key: cord-0004782-zhrjo9w3
authors: nan
title: SPR 2019
date: 2019-04-01
journal: Pediatr Radiol
DOI: 10.1007/s00247-019-04376-7
sha: c80b6ec28f78c81fd30ef5e1c7ca436fd9c78b15
doc_id: 4782
cord_uid: zhrjo9w3

nan

Cristian Garcia, MD 2009 Antônio Soares de Souza, Stephen Chapman, MD

This SPR Presidential Recognition is awarded to Patricia E. Burrows. This award recognizes her fantastic career and life-long dedication to the image-guided treatment of patients, in particular, with vascular malformations and hemagiomas, and patients with cardiovascular and neurovascular diseases. Besides numerous patients whom she has positively altered their lives, our Society and our membership have also benefited from all her teachings at past SPR meetings and her numerous academic publications. Her clinical and research works epitomize the mission of SPR: to foster excellence in pediatric health care through imaging and image-guided care.

Pat grew up in southern Manitoba, Canada, and received her MD at the University of Manitoba Faculty of Medicine in 1976. She started in a pediatric residency but switched to radiology after 1 year and completed her residency in Winnipeg, Manitoba in 1981. She completed a fellowship in pediatric radiology at Winnipeg Children's Hospital, under the mentorship of Martin Reed and Rod McPherson. She enjoyed doing extra angiography during her residency, as well as a rotation in the cardiac Cath Lab at Children's Hospital. This led to a second fellowship in what was called "pediatric special procedures" including congenital heart disease, angiography, and neuroradiology at Children's Hospital Boston, Harvard Medical School in 1982. She was assigned a research project correlating angiographic findings with John Mulliken's new classification of vascular anomalies.

After her fellowship, she moved with a group of cardiovascular surgeons to Arkansas Children's Hospital in Little Rock. This was shortlived because of her immigration status and she took a position at The Hospital for Sick Children's Hospital in Toronto in 1983, as head of cardiac radiology. She performed pediatric cardiac catheterizations and cardiac interventions and developed pediatric cardiac MRI locally. She continued practicing general pediatric radiology and angiography and then became involved in pediatric neuroimaging. While in Toronto, she met Pierre Lasjaunias, a French neuro-interventionalist who was recruited to train the neuroradiologists at the Toronto Western Hospital in neuro-interventional techniques. Lasjaunias agreed to train Pat in Paris and she spent seven months with him in 1985 to 86 where she learnt learned even more about vascular anomalies, neurovascular anatomy and interventional techniques. In this pre-micro catheter era, she modified the adult techniques for the pediatric patients including embolization of arteriovenous fistula using detachable balloons, coils and glue, as well as ethanol and sodium tetradecyl sulfate sclerotherapy for venous and lymphatic malformations. She also had the opportunity to work with Dr. Colapinto and subsequently published the first series of pediatric trans-jugular liver biopsies and the first case report of embolization for pediatric traumatic priapism. She was also the first person to use platinum fiber micro coils in Canada and the first to perform renal artery stenting (with Chet Rees) in a child.

In 1994, Pat was recruited to Boston Children's Hospital as chief of Pediatric Interventional Radiology. With this move, she decided to give up cardiac imaging in order to focus on vascular anomalies. With John Mulliken, plastic surgeon, and Judah Folkman, vascular scientist, Pat was a cofounder and subsequent codirector of the Vascular Anomalies Center [VAC] at Boston Children's Hospital. She realized very early on the importance of a multidisciplinary team for optimizing both the work-up and the therapeutic alternatives in the field of vascular congenital anomalies. This VAC became the model for many other institutions for the treatment of patients with vascular anomalies. Her motto in congenital vascular anomalies was straight forward "you cannot treat if you don't understand the lesion".

She became a member of the International Society for the Study of Vascular Anomalies [ISSVA] in 1994, and served on the Board of Directors from 2006 to 2014, and as president from 2010 to 2012. While in Boston, Pat began using doxycycline to treat cystic lymphatic malformations and subsequently showed that it was safe and effective to use in infants. She trained many fellows and residents, and two of them, Ahmad Alomari and David Lord, went on to direct Vascular Anomalies programs themselves. Radiology based postdoctoral fellowship in vascular anomalies research under the direction of Dr. Ramani Ramchandran, head of developmental vascular biology. This program resulted in new observations of the role of endothelial cilia in vasculogenesis which were recently published, as well as a zebrafish model for AVM being used for drug screening. Pat retired on August 31, 2018 and has relocated to St. Petersburg, Florida. Upon retirement, Pat received Congressional tribute for all her contributions in advancement and discovery of new knowledge in the field of vascular malformations from the 8th Congressional district of Michigan. Besides numerous visiting professorships and keynote addresses which she cherishes, Pat is most proud of her many key publications in vascular anomalies and also of her namesake for the chair in Pediatric Interventional Radiology and Vascular Anomalies at Children's Hospital Boston.

She enjoys spending her time with her spouse, Jay, sharing their passion for sailing. When she is on land, she enjoys gardening and has become an excellent potter in the past 7 years! From handling and steering that tracker catheter anywhere in all large and small vessels from head to toe, Pat is bringing her concentration and dexterity to pottery and may all potters beware!!!

Dr. John P. Caffey was regarded throughout the world as the father of pediatric radiology. His classic textbook, "Pediatric X-Ray Diagnosis", which was first published in 1945, has become the recognized bible and authority in its field. The seventh edition of this book was completed several months before his death in 1978. It has been among the most successful books of its kind in the medical field.

Dr. Caffey was born in Castle Gate, Utah on March 30, 1895. It is interesting that he was born in the same year that Roentgen discovered the x-ray. Dr. Caffey was graduated from University of Michigan Medical School in 1919, following which he served an internship in internal medicine at Barnes Hospital in St. Louis. He spent three years in Eastern Europe with the American Red Cross and the American Relief Administration, and returned to the United States for additional training in medicine and in pediatrics at the Universities of Michigan and Columbia, respectively.

While in the private practice of pediatrics in New York City at the old Babies Hospital of Columbia University College of Physicians and Surgeons, he become interested in radiology and was charged with developing a department of pediatric radiology in 1929. He frequently expressed appreciation and admiration for Ross Golden, Chairman of Radiology at Columbia Presbyterian Hospital, who allowed him to develop a separate department of diagnostic radiology without undue interference, and who was always available to help and advise him.

Dr. Caffey's keen intelligence and inquiring mind quickly established him as the leader in the fields of pediatric x-ray diagnosis, which recognition became worldwide almost instantaneously with the publication of his book in 1945.

Dr. Caffey received many awards in recognition of his achievements. Outstanding among these were the Dr. Caffey's contributions to the pediatric radiologic literature were many. He was instrumental in directing attention to the fact that a prominent thymic shadow was a sign of good health and not of disease, an observation that literally spelled the end to the practice of thymic irradiation in infancy. Infantile cortical hyperostosis was described by him and is called "Caffey's Disease". Dr. Caffey in 1946 first recognized the telltale radiographic changes that characterize the battered child, and his students helped disseminate his teachings about these findings. It was Dr. Caffey who first recognized and descried the characteristic bony changes in vitamin A poisoning. He recognized and described the findings associated with prenatal bowing of the skeleton.

In 1963, 3 years after his retirement from Babies Hospital, he joined the staff of the Children's Hospital of Pittsburgh as associate radiologist and as Visiting Professor of Radiology and Pediatrics at the University of Pittsburgh School of Medicine. Although Dr. Caffey came to Children's Hospital and the University of Pittsburgh in an emeritus position, he worked daily and on weekends throughout the years he was there. In Pittsburgh, he made four major new contributions to the medical literature. He described the entity, "idiopathic familial hyperphosphatasemia". He recognized and described the earliest radiological changes in Perthes' Disease. He called attention to the potentially serious effects of shaking children, and used this as a subject of his Jacobi Award lecture. He described, with the late Dr. Kenny, a hitherto unrecognized form of dwarfism, which is now known as the Caffey-Kenny dwarf. The John Caffey Society, which includes as its members pediatric radiologists who have been intimately associated with Dr. Caffey, or who have been trained by his students, was established in 1961. This society is now among the most prestigious in the field of radiology. His book and the society named in his honor will live on as important memorials to this great man.

His greatness was obvious to all who worked with him. He was warm, kind, stimulating, argumentative, and above all, honest in his approach to medicine and to x-ray diagnoses. His dedication to the truth was expressed in his abiding interest in the limitations of x-ray signs in pediatric diagnosis and in his interest in normal variation in the growing skeleton. He was concerned with the written and spoken word and was a skilled semanticist. His book and his articles are masterpieces of language and construction. He stimulated and was stimulated and loved by all who had the privilege of working with him. Radiology and Pediatrics have lost a great man, but they shall ever have been enriched by his presence.

Bertram R. Girdany, MD JOHN P. CAFFEY AWARD PAPERS 

Authors are listed in the order provided. An author listed in bold identifies the presenting author. Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Assessing the added value of contrast enhanced US in detecting the vascular complications of liver transplantation in children.

We retrospectively evaluated all Doppler US and contrast enhanced ultrasound (ceUS) performed since 8/2017 when we started to routinely use ceUS.

For the first 5 post-operative days (POD) All patients had twice a day Doppler US. ceUS was performed as a baseline in the first POD and to confirm any suspected vascular complications. Demographic, clinical and imaging findings were recorded and were correlated with surgery. Results: 14 children (9 females) had 15 liver transplants during this period. The mean age was 4.1 years old (range 6 months-16.1 years). Indications of the liver transplants were as follows; biliary atresia (n=9), alpha 1 antityripsin deficiency (n=1), metabolic liver disease (n=2), cystic fibrosis (n=1) and cryptogenic cirrhosis (n=1). Three patients had split liver and 11 patients had whole liver transplantation.ceUS was successfully performed in all patients. There were no complications related to the ceUS. In three patients CEUS was performed to evaluate for vascular complications following abnormal Doppler US. In two patients ceUS demonstrated vascular thrombosis confirmed in surgery: In a six month old male post whole liver transplant, initial ceUS in POD 1, confirmed portal vein thrombosis (PVT) seen on Doppler US. Surgery confirmed PVT but also hepatic artery (HA) thrombosis. Marked HA narrowing in the initial ceUS could be seen retrospectively when compared with a post thrombectomy follow-up ceUS. In a one year old female post whole liver transplant, initial ceUS was normal. On POD 3, Doppler showed tardus parvus flow in intrahepatic arteries this followed immediately by ceUS that demonstrated marked HA stenosis. This led to seven hours follow-up ceUS that showed progression to complete HA occlusion confirmed in surgery. In a 16 year-old female, POD 1 ceUS excluded right hepatic vein thrombosis suggested by Doppler and Color US.

Conclusions: Our pilot study shows that ceUS can provide definite bed side diagnosis of vascular complications in the immediate post liver transplant period. Use of ceUS has the potential to provide earlier diagnosis of liver vascular complications and obviate the need to perform contrast enhanced CT and reduce unnecessary exploratory surgeries. Purpose or Case Report: Gadoxetate disodium (Eovist ®) is the preferred MR contrast agent for pediatric hepatobiliary imaging. It is known to cause arterial phase transient severe breathing motion artifacts in 5-22% of adults, and adult studies have raised caution against its use for evaluation of arterial enhancing lesions.This study seeks to evaluate the frequency of transient severe respiratory motion secondary to gadoxetate disodium in a pediatric cohort.

This is a retrospective, IRB approved study with informed consent waiver. The radiology information system of a children's hospital was searched to identify all MRI studies performed with gadoxetate disodium during January 2016-June 2018. Exclusion criterion was MRI studies with incomplete set of dynamic post contrast imaging. Two readers independently evaluated the pre-contrast, arterial, portal and equilibrium phase of dynamic liver imaging for respiratory motion artifact on a 5 point scale (1 no motion, 2 mild motion, 3 moderate motion, 4 severe motion-still diagnostic, 5 extreme motion-not diagnostic). Average scores of the 2 readers for each phase were used for analysis. Transient severe respiratory motion was defined as an increase in breathing motion score of ≥ 1.5 from pre contrast to arterial phase that returned to < 3 in equilibrium phase of imaging.

The study cohort consisted of 140 cases (60% female), age range: 1 month -23 years (median 13 years). 102/140 scans were performed non-sedated. 98 scans were performed on 1.5T scanners and 42 on 3T scanners. Mean respiratory motion score for each phase of scan for the entire cohort was: pre-contrast: 2.23, arterial: 2.56, portal venous: 2.39, equilibrium: 2.31. Breathing motion score increased by a factor of ≥ 1.5 in 15/140 (10.71%) cases after injection of gadoxetate disodium. The score remained at ≥ 3 in 7/15 cases, and returned to < 3 in 8 cases resulting in a transient severe respiratory motion rate of 5.71 % (8/140 cases).

The rate of transient severe respiratory motion artifacts in the pediatric population is estimated at 5.7%, similar to reported rates in the adult population. Pediatric radiologists should be aware of this potential limitation of gadoxetate disodium when performing hepatobiliary MRI in children.

Purpose or Case Report: To compare hepatic proton density fat fraction (PDFF), R2* values, and respiratory motion between navigator-gated and breath-held multi-echo Dixon (mDixon) magnetic resonance imaging (MRI) acquisition techniques in children and young adults with suspected liver disease.

This retrospective study was approved by the institutional review board with a waiver of informed consent. Patients who underwent liver MRI with breath-held and navigator-gated mDixon sequences between January 2017 and July 2018 were included. One reviewer, blinded to sequence, placed free-hand regions of interest, inclusive of as much liver as possible, on four images from each PDFF and R2* parametric map. Another blinded reviewer graded respiratory motion using a 5-point Likert scale. Pearson correlation (r), Lin's concordance coefficients (rc), and Bland-Altman analyses were used to assess agreement between techniques. Frequency of clinically-liming motion (score≥3) was compared with Fisher's exact test. Results: 42 patients were included (15F:27M; mean age:15.7±4.6 years). Mean PDFF and R2* were 16.6±13.1% and 29.3±4.7 sec -1 on breath-held images vs. 17.0±13.2% and 29.6±5.2 sec -1 on navigator-gated images. PDFF demonstrated almost perfect agreement between sequences (rc=0.997, 95%CI:0.994-0.998; mean bias:0.3%; 95% limits of agreement:-2.4 to +1.7%), while R2* values demonstrated very strong correlation but poor agreement (r=0.837, rc=0.832, 95%CI:0.716-0.910). Navigator-gated images exhibited significantly higher frequency of clinically-limiting respiratory motion (88% vs. 48%, p=0.0001). Conclusions: Despite greater respiratory motion artifact, a freebreathing navigator-gated mDixon sequence produces PDFF values with almost perfect agreement to a breath-held sequence and thus may be an option in patients with limited breathholding capacity. Purpose or Case Report: Applying lower diffusion gradient eliminates signal from all the vessels thereby improving visibility of T2-hyperintense lesions on image with low b-value of 50-100 s/mm2 (LBV). LBV have been shown to be superior to T2-weighted fast spin echo sequence (T2W) in detection of liver lesions in adults. There are no such studies assessing this in children. The purpose of the study is to compare the sensitivity of LBV images and T2W images in the detection of focal liver lesions in children.

A retrospective review of liver MRI performed for assessment of focal liver lesions in 50 children (22 males, 28 females; age 2 months to 17 years (mean 10.88 years)) was done. Two radiologists reviewed both LBV and T2W sequences independently at different occasions to note number of lesions, smallest lesion size, and location. A consensus reading of the entire MRI examination and correlation with follow-up, other imaging modalities, and pathology in available cases, was used to determine final number of lesions as a reference standard. Inter-observer agreement between 2 sequences for each reader for detection of number of lesions was assessed using intra-class correlation coefficient (ICC). The average number of lesions per patient detected by both readers on each sequence was compared with each other and with the reference standard using ICC and Signed-Rank test. The smallest lesions detected by each sequence were compared using paired t-test. Results: A total of 170 hyperintense lesions were identified on consensus review to serve as reference standard. The average number of lesions identified by both readers on LBV was 134 (79%) and on T2W were 95 (56%). There was excellent interobserver agreement for detection of lesions on LBV (ICC=0.96 (0.93-0.98)) and T2W (ICC=0.85 (0.75-0.91)), with slightly better agreement on LBV. Both readers identified significantly more lesions on LBV compared to T2W (reader 1 p=0.0036, reader 2 p=0.0001). Compared to the reference standard (mean number of lesions=3.45), T2W detected significantly less lesions (mean number of lesions=1.91; p=0.0000) while there was no significant difference in lesion detection on LBV (mean number of lesions-2.68; p=0.1527). LBV and T2W were not significantly different in identifying the smallest lesion size (reader 1 p=0.50, reader 2 p=0.53). Conclusions: Low b-value DWI images are more sensitive than T2-weighted sequences in detecting hyperintense focal liver lesions in children. Purpose or Case Report: Pancreatic atrophy, a finding of chronic pancreatitis, has traditionally been defined qualitatively. Volumetric measurements of the pancreas may provide a means to define atrophy more quantitatively but an understanding of normal volume is required. The purpose of this study was to define normal pancreatic parenchymal volume in children, its relationship with demographic and anthropometric factors and its relationship to volume of fluid secreted after secretin administration.

This study was IRB-approved and compliant with HIPAA. A single observer manually segmented (Vitrea, Vital Images) the pancreas on axial MR images prospectively obtained in 50 children (6-16 years) without pancreatic disease. Parenchymal volumes were correlated (Pearson) with age, height, weight, body surface area (BSA) and with previously calculated volume of fluid secreted in response to secretin. Student's t-test was used to compare means. Quantile regression was used to define 5 th /95 th percentiles for parenchymal volume by BSA. Results: Mean (±SD) parenchymal volume for our 50 healthy children was 46.0±18.8mL. There was no significant difference in parenchymal volume by sex. There were statistically significant correlations between parenchymal volume and age (r = 0.51, p=0.002), height (r=0.67, p<_0.0001), weight (r=0.75, p<0.0001) and BSA (r=0.75, p<0.0001). The 5 th percentile for parenchymal volume by BSA could be calculated by: = -4.97+24.66xBSA.Parenchymal volume was statistically significantly correlated with secreted fluid volume in response to secretin (r=0.51, p=0.0002).

In this study we have defined normal pancreatic volumes for children, and shown that these values vary with patient age and size. We have also defined the 5 th percentile for parenchymal volume by BSA, below which atrophy might be considered to be present as a feature of chronic pancreatitis. We have also shown that pancreatic parenchymal volume was moderately correlated with volume of fluid secreted by the pancreas in response to secretin. Purpose or Case Report: Normal values for linear thickness of the pancreas in children have recently been established on CT, but have not been validated for MRI. Additionally, a normal T1 ratio (pancreas:spleen) has been defined in adults but this has not been validated in children. The purposes of this study were: 1) To validate previously defined normal pancreatic thickness measurements for MRI and in a distinct pediatric population. 2) To validate the previously defined T1 ratio (pancreas:spleen) for adults in a pediatric population and assess its relationship to demographic factors.

This study was IRB approved and HIPAA compliant. A single observer measured linear thickness of the pancreas in the four previously defined segments (head, neck, body and tail) on axial images from 50 MRI exams prospectively obtained in children (6-16 years old) without pancreatic disease. ROIs were also placed in each the pancreas and spleen (on the same slice, if possible) on T1-weighted gradient recalled echo (GRE) images to measure the T1 ratio. Linear measurements were compared to previously defined norms. T1 ratios were summarized with means and standard deviations and assessed for their relationship to age, sex, height, weight and body surface area (BSA) using t-tests and linear regression, as appropriate. Results: Using previously defined normal values for pediatric pancreatic thickness, 34 healthy participants (68%) had "atrophy" of one segment and 14 (28%) had "atrophy" of two or more segments.Mean (±SD) T1 ratio for the study population was 1.33±0.15. There was a statistically significant correlation between T1 ratio and age and height (r=-0.44, p=0.0015 and r=-0.38, p=0.0069, respectively). but not with weight or BSA. There was no significant difference in T1 ratio based on sex. 11 participants (22.4%) had a T1 signal ratio <1.2, a threshold previously determined to correspond with exocrine insufficiency in an adult population. Conclusions: We demonstrate that previously defined normal pancreas thickness values for children are likely too restrictive and that better measures of pancreas bulk in children are needed. Mean T1 ratio in children is related to age and height and is above the normal threshold of 1.2 previously described in adults but >20% of children have T1 ratios below the adult cutoff suggestive of exocrine insufficiency. Specific cut-offs for disease identification by T1 ratio in children are needed. Purpose or Case Report: Magnetic resonance imaging (MRI) T1 relaxation time is altered by fibrosis and inflammation and is a potential marker for diseases of the solid abdominal organs. The purpose of this study was to measure T1 values of the liver, pancreas, and spleen in healthy children.

This cross-sectional IRB-approved study prospectively recruited healthy children aged 7 to 17 years with BMI in the 5 th to 85 th percentile to undergo abdominal MRI at 1.5 or 3T, including T1 mapping with a variant Modified Look-Locker (MOLLI) sequence between February 2018 and August 2018. A single reviewer placed free-hand regions of interest on the T1 parametric maps in the liver, pancreas, and spleen, inclusive of as much parenchyma as possible and on up to four axial images per organ. Student t-tests were used to identify differences in T1 values by gender. Linear regression was performed to assess associations between T1 values and age. Results: 32 patients were included in the study (16F:16M; mean age:12.2±3.1 years; n=16 at 1.5T; n=16 at 3T). Median T1 relaxation times per organ at 1.5T were: 1) liver: 569±39 ms; 2) pancreas: 576±55 ms; and 3) spleen: 1172±71 ms and at 3T were 1) liver: 767±63 ms; 2) pancreas: 730±30 ms; and 3) spleen: 1356±87 ms. T1 values were not significantly different between males and females at either field strength. Linear regression showed no significant association between age and T1 values of the liver, pancreas, and spleen at 1. 5T (r=0.66, p=0.30; r=0.48, p=0.094; and r=0.49, p=0.057; respectively) and 3T (r=0.50, p=0.12; r=0.56, p=0.15; and r=0.40, p=0.12; respectively) . Conclusions: We report normal MRI T1 values for the liver, pancreas, and spleen at 1.5 and 3T in a cohort of healthy children and observed no significant association between age or sex and T1 values. While these values might be used to exclude disease, a larger cohort will likely be needed to establish cut-off values to allow future comparisons with disease states. Purpose or Case Report: Percutaneous needle placement relies on image guidance using X-ray, CT, US, or MRI. MRI-guided interventions are ideal for pediatric patients to eliminate ionizing radiation while providing superior imaging of the nervous and musculoskeletal systems. However, performing procedures within a closed-bore MRI scanner is challenging because of the ergonomics of limited patient access. We are developing small, MRI-compatible robotic systems that will enable MRI-guided interventions.

Methods & Materials: Three systems are being developed for bone biopsy, arthrography, and nerve block/ablation. For bone biopsy, a table-mounted robot with bone drill guide was developed and tested in a cadaveric leg to evaluate feasibility and accuracy. For arthrography, a patient-mounted four degree of freedom robotic positioning and orienting stage was developed and tested in cadaveric shoulder and hip joints. For nerve block/ablation, a two degree of freedom needle driving and rotation stage will allow remote needle advancement under real-time MRI guidance. A software interface allows the physician to select the target and skin entry point which define the safest needle trajectory.

The bone biopsy robot study shows feasibility and clinically acceptable bone drill accuracy in a cadaver leg ( Figure  1 ). A total of 10 biopsy targets were sampled using MRI guidance, 5 from the femur and 5 from the tibia. All of the targets were successfully reached with an average targeting accuracy of 1.43 mm. The arthrography robot study in Thiel embalmed cadavers shows feasibility and good success for intra-articular injection during MR arthrography ( Figure 2 ). In one male and one female cadaver, a total of 13 roboticallytargeted joint injections were performed (10 shoulder, 3 hip). All were successful with an average procedure time of 20 minutes. The next step is a pilot clinical trial.

The robotic devices presented here are being developed to enable radiation-free MRI-guided procedures. Clinical application in bone biopsy of suspicious lesions, joint injection for MR arthrography, and nerve block/ablation for pain management are pertinent to pediatric patients. The patientmounted design minimizes challenges of patient motion and MR-compatibility allows safe use in the scanner bore without degradation of image quality. These devices have the potential to enable physicians to perform MR interventions with high accuracy and safety and also shorten procedure duration.

Purpose or Case Report: To investigate the long-term MRI imaging findings and clinical outcomes of pediatric patients with aneurysmal bone cysts treated by sclerotherapy at a large tertiary children's hospital.

An IRB-approved retrospective review was performed on patients who completed sclerotherapy for aneurysmal bone cyst from 2007 through 2018. Patients were excluded if they had no pre-treatment MRI, post-treatment MRI, or received surgery between sclerotherapy treatments, yielding a total number of 38 patients. Pre-treatment and final posttreatment MRI imaging and reports were reviewed. The sclerotherapy treatment(s) and surgical history were recorded.

Electronic medical records were also reviewed for clinical symptoms prior to initial presentation and preceding the final MRI study. The data analysis for this study was generated using SAS software. Conclusions: There is little information about the evolution of MRI imaging findings following successful sclerotherapy of aneurysmal bone cysts. This study provides a summary of common MRI findings in aneurysmal bone cysts and clinical outcomes for these patients after sclerotherapy. Purpose or Case Report: To confirm technical feasibility and clinical efficacy of osteoid osteoma (OO) cryoablation in a large, pediatric/adolescent cohort with short and long term follow-up. Methods & Materials: An electronic medical record and imaging archive review was performed to identify all cryoablations performed for OOs between 2011 and 2018 at a single tertiary care pediatric hospital. The analysis included 63 patients with suspected OOs treated by cryoablation (age range, 3-18 y; mean age, 11.7 y; 37 boys; 26 girls). Conventional CT guidance was used in 22 procedures; cone-beam CT needle guidance was used in the remaining procedures. Follow-up data were obtained via a standardized telephone questionnaire and clinical notes. Results: Technical success, defined as placement of cryoprobes in the desired location, was achieved in 100% of the 63 patients.Immediate clinical success (cessation of pain and nonsteroidal antiinflammatory drug [NSAID] use within 1 mo after the procedure) was achieved in 61/63 (96.8%) of patients. There were 2/63 (3.2%) clinical failures, both in small bones of the feet. 5/63 (7.9%) patients had clinical recurrence with repeat ablation and subsequent complete response. Furthermore, longterm clinical success (cessation of pain and NSAID use for > 12 mo after the procedure) was achieved in 54/57 patients (94.7%) at the time of this submission. There were no major complications. Two patients (3.2%) were admitted overnight for pain control after the procedure (minor complications). There were no other minor complications. Conclusions: Image-guided cryoablation is a known technically feasible, clinically efficacious, and safe therapeutic option for children and adolescents with symptomatic OO. Long-term results in this study provide confirmation of a definitive treatment paradigm. Purpose or Case Report: Biopsy is often performed to aid in diagnosis and to clarify optimum management of children with suspected osteomyelitis. We describe our experience with bone biopsy for pediatric osteomyelitis and analyze yield with regard to imaging features on MRI and procedure technique.

This was a retrospective review of patients with suspected osteomyelitis who underwent percutaneous bone biopsy in interventional radiology. Review parameters included MRI features, technical aspects of the procedure, and clinical data, including details of clinical presentation and microbiology and pathology results. Results: Forty patients (mean age 9.5y, range 0.4-25.8y) underwent biopsy. Fluoroscopy was used in 29 cases. CT guidance was used in 9 and US guidance was used in 2. Additional US guidance for adjacent soft tissue biopsy was used in 15 cases. A 16G bone biopsy cannula was utilized in 34 cases; 13G in 5; and an 11G in 1 case. The mean number of bone cores obtained was 3.6 (range 1-10). There were no procedure-related complications.Biopsy culture yielded a pathogen in 4 of 40 cases, and culture yield did not show significant correlation with fever, leukocytosis/blood culture, ESR/CRP, number of cores, instrument gauge, or acquisition of liquid aspirate or soft tissue. Based on culture and/or pathology results, 23 patients were diagnosed with infectious osteomyelitis (17 acute and 6 chronic). One patient was diagnosed with a stress fracture and one with Hodgkin lymphoma. Chronic recurrent multifocal osteomyelitis (CRMO) was diagnosed in 15 patients; CRMO was supported on pathology in 9 cases. Biopsy results, therefore, did not directly inform diagnosis in 6 (15%) patients.On MRI, 74% of lesions were in the appendicular skeleton and 26% in the axial skeleton. An axial location demonstrated a significant association with a positive bacterial culture (p=0.04). There was also a statistically significant association between a lesion size ≤20 mm on MRI and negative bacterial culture (p<0.01). There was no association between culture yield with interosseous/subperiosteal fluid or myositis. Conclusions: Results support bone biopsy as a safe and effective procedure in the management of pediatric osteomyelitis. Although culture yield from bone biopsy was low, overall biopsy results aided in diagnosis in a majority (85%) of patients with suspected osteomyelitis. On MRI, lesion size ≥20 mm and an axial location demonstrated association with positive culture. Purpose or Case Report: Adequate gastric distension via insufflation is a key step in creating a safe percutaneous window during gastrostomy/gastrojejunostomy (G/GJ) placement. However, this may be limited due to rapid egress of air from the stomach into the duodenum, despite the use of glucagon. Herein, we describe an adjunctive novel technique of transgastric balloon occlusion to maximize gastric insufflation and assess the outcome of this technique during G/GJ placement in children.

A single-center, IRB-approved, retrospective review of 15 patients (6 female and 10 males) with a mean age of 4.8±5.6 years (range 0-16 years) and mean weight of 20.9±18.9 kg (range 3-54.2 kg) who had G/GJ placement utilizing transgastric balloon occlusion over a 2 year period. The standard technique was antegrade placement with administration of glucagon. After initial percutaneous failure, more recently the practice of positioning a balloon in the proximal duodenum was adopted to temporarily obstruct the gastric outlet. Clinical history, patient demographics, procedure reports, balloon type, technical success, and outcomes were reviewed.

Results: The addition of a transgastric balloon occlusion was successful in salvaging G/GJ tube placement in 10/15 (67%) patients (3 G, 7 GJ) which likely would have been unsuccessful using standard practice. Most common underlying disorders included cardiac (5), neurologic (3), and oncologic (2). Of 5 unsuccessful placements, 3 were attributed to persistent colonic interposition/high position, 2 to high stomach position. Subsequently, 4 of 5 underwent surgical gastrostomy placement. No procedure-related complications were reported.

The technical success of G/GJ placement in children with challenging percutaneous access may be improved by the novel use of transgastric balloon occlusion to prevent rapid egress of air out of the stomach. A larger cohort is needed to assess the technical efficacy and to identify ideal balloon characteristics and insufflation location based on patient and disease characteristics. Additional challenges to the workflow included the addition of a third location for daily consultation rounds at the children's hospital NICU and the elimination of a campus shuttle route. While the pediatric radiology department's goal was to maintain the high level of service, availability, and communication with the neonatology team, concerns regarding time and movement energy waste due to travel threatened to make this untenable.

A reliable process for conducting patient care rounds with participants physically located in different institutions across the medical campus was established using a web-based teleconferencing tool. Requirements of the application were that connectivity should be (nearly) instantaneous across the networks of three different hospitals, reliable, and user-friendly, with two-way audio and video, including screen sharing to facility discussion of interesting radiology findings. Schedule, daily electronic meeting apointments were sent to the assigned pediatric radiologist and the NICU conference room at the UH, where clinical team rounds occurred over a two-hour period daily. On-site rounds were preserved at the two larger NICUs which were also in closer proximity to one another. Results: Travel times were significantly reduced following implementation of remote consultation at the UH NICU facility, with an estimated savings of 4940 minutes, 113.1 miles, and 225,420 steps over a six month timeframe. Feedback from pediatric radiology faculty and NICU treatment teams was widely positive following the implementation of telerounds. We demonstrated the ability to provide a similar level of quality of communication, ability to entertain dialogue regarding exams, and timeliness of rounds and a significant increase in satisfaction after implementation across all metrics.

The implementation of tele-radiology NICU rounds resulted in savings in time and travel and improved satisfaction with communication for both radiology and NICU providers.

Implementing the "What-Where-When" approach to improve patient history availability at the time of radiograph interpretation CPT concern for abuse changed in 8 cases (6.2%; 95% CI 3.1,12.0), increased in 5 and decreased in 3. After consideration of the constellation of clinical findings, CPT concern rarely changed from abuse to non-abuse (N=1) or from non-abuse to abuse (N=1). CPT medical management changed in 11.6% (95% CI 7.1, 18.5). CPT reported that the second read process was helpful in increasing confidence in SS interpretation in over 90% of cases. Conclusions: Our pilot identified agreement in fracture presence, absence, or certainty in >90% of cases. The second read process resulted in a change in the level of concern for abuse in <10% of cases. Additional data are needed to understand if certain SS findings are at higher risk for disagreement to understand the utility of targeted second read programs. Purpose or Case Report: There has been recent increasing interest in imaging protocol standardization. At our institution, challenges to standardization stem from being a large hospital system with multiple sites and 10 MRI scanners at our facility alone. Scanners are different ages, designed by different manufacturers and vary by type. Posted protocols were not standardized by name and would often not match those within the MRI scanners. Finally, sequence names on MRI scanners varied by manufacturer and type of scanner, and these names would not match posted protocol names. We aim to optimize imaging protocols by standardizing the MRI naming system, making protocols readily available, and continuing to maintain image quality. Our goal is to avoid communication breakdown between technologists, radiologists and referring physicians and to add value by providing consistency.

A multidisciplinary team including technologists, radiologists and physicists was assembled and a standard sequence naming system was agreed upon. The process of updating the online posted protocols was re-evaluated and representative "thumbnail" images were associated with each sequence in a protocol to avoid confusion. A soft launch of the new system was implemented with the most frequently used MRI MSK protocol in September 2018. An electronic survey was distributed in October to identify challenges and acceptance by technologists. Results: 22 MRI technologists completed the survey. Of all respondents, 95.5% (n=21) found the new naming convention easier to understand with 81% also adding that it positively affected their workflow (n=18). 57.1% of respondents utilized the new protocol (n=13) since the soft launch the month prior. Of those who utilized it, 100% found the new thumbnail images helpful during the workday (n=13), and 100% also noted that it should be adopted for all protocols (n=13). 90.9% of responding technologists noted they would refer to the posted protocols more often if thumbnail images were included in all protocols (n=20). Individual suggestions for improvement included indicating anatomic coverage on thumbnail photos, including unusual scanning angles, and demonstrating preferred patient positioning. Conclusions: MRI protocol optimization including standardization of protocols, sequences, coverage and a reference article supporting protocol design was well received by technologists and improves their daily workflow. We are also working with ordering clinics, in an effort to divert more patients to unsedated scans. We are diligently working to shorten protocols and sequencing times for scans to further reduce the SMRI burden. This future abbreviated approach, with the aid of VR scans, will allow for our Child Life team to further stratify patients towards non-sedate slots. Purpose or Case Report: The purpose of our study was to assess the accuracy of measuring glomerular filtration rate (GFR) using motion-robust high spatiotemporal resolution 3D dynamic contrast enhanced MRI, by comparison of MRI-GFR to GFR from 99mTcDTPA nuclear medicine study (NM-GFR).

This IRB approved study enrolled children, between 0-20 years, undergoing both a clinically indicated contrast enhanced MRI, and a nuclear medicine GFR study within 2 weeks of each other. Each patient consented for the study underwent an additional 6-minute dynamic contrast enhanced MRI scan using the motion-robust high spatiotemporal resolution dynamic radial VIBE sequence (Siemens 3T). The images were reconstructed offline using compressed sensing image reconstruction including regularization in temporal dimension to improve image quality and reduce streaking artifacts due to fast imaging. Images were then automatically post processed using in-house developed software. Post processing steps included segmentation of kidney parenchyma and aorta using a convolutional neural network technique and tracer kinetic model fitting using the Sourbron's two-compartment tracer kinetic model to calculate the magnetic resonance based GFR (MRI-GFR). The MRI-GFR results were compared to the glomerular filtration rate measured by 99mTcDTPA nuclear medicine study (NM-GFR). Results: 18 children (average age 5.9, 8 female:10 male) were enrolled between July 2017-present. The method was technically feasible in all patients. The results showed that the MRI-GFR correlated with NM-GFR with r-square value of 0.94. We also performed Bland Altman analysis (i.e. difference of MRI-GFR and NM-GFR versus mean of NM-GFR and MR-GFR), which showed a reproducibility coefficient (RPC) of 22 with 95% confidence interval, and the coefficient of variation (CV) of 8.2% with values between -21 (-1.96 standard deviation) and 24 (+1.96 standard deviation). Conclusions: MR-GFR is a reliable method of measuring glomerular filtration rate and is comparable to the GFR measured by plasma clearance of 99mTcDTPA. 

Our results provide evidence that IVIM imaging is useful to evaluate split renal function. We propose that the separation of perfusion and diffusion processes is necessary to estimate renal function in children using diffusion MRI. Image processing and statistical models adjusting for patient age were necessary to reveal correlations between microvascular perfusion and function, however, we did not find similar correlation with the diffusion coefficient. Purpose or Case Report: To report the frequency of calyceal diverticula (CD) as seen at functional MR urography (fMRU) and determine the ability of fMRU to diagnose calyceal diverticulum in a pediatric cohort.

This is an IRB-approved retrospective study of all patients with suspected CD that underwent fMRU at our institution. Two-pediatric radiologists reviewed each fMRU, blinded to clinical information and other imaging, to determine the presence, morphology and contrast accumulation in cystic lesions suspected of being CD. Each radiologist made a determination of cyst or CD solely based on the presence of contrast within the lesion. Size and location were also recorded. The timing at which the contrast was first seen within the lesion and the pattern of contrast within the lesion (i.e. complete filling versus layering contrast) were also recorded. The diagnosis of CD was confirmed by either surgery, retrograde pyelogram or acknowledgement of fMRU results in the urologic clinical note. Chi-square was used to examine differences between characteristics of cysts versus diverticula. Inter-reader agreement was also calculated with kappa. Purpose or Case Report: Generalized arterial calcification of infancy (GACI), also known as idiopathic infantile arterial calcification, is a rare genetic disorder characterized by calcifications and injury to large and medium size vessels. We aim to describe the imaging findings of GACI in children.

This is a retrospective review of initial and follow up CT, CTA, MRA, and vascular ultrasound imaging in children with confirmed GACI at a single institution. All initial imaging studies were reviewed for the presence and distribution of arterial calcifications, stenosis, and wall thickening/irregularity within the chest, abdomen, and pelvis. Available follow up studies were compared to initial imaging findings. A chart review was performed for clinical history. Results: A total of 8 patients (5 boys) from 6 families (2 sets of siblings) were identified. Patients presented prenatally (n=4); or with cardiac failure (n=2), seizures (n=1), or hypertension (n=1). Average follow up was 37 months (range 0 -11 years). A single death occurred at 2 months of age.5 patients underwent imaging at birth and all had arterial calcifications as follows: aorta (n=4), pulmonary (n=3), mesenteric (n=3), renal (n=3), iliac (n=3), subclavian (n=2), brachiocephalic (n=1), common carotid (n=1), coronary (n=1), and splenic (n=1) arteries. 1 patient had main pulmonary artery stenosis and 1 had bilateral renal artery stenosis.4 of the newborn patients underwent follow up imaging between 1 month and 3 years of age. Calcifications were: decreased (n=2), resolved (n=1), or stable (n=1). Stenosis was new or progressed in these 4 patients as follows: renal (n=3), pulmonary (n=2), aorta (n=2), mesenteric (n=2), and iliac (n=1) arteries. Vessel wall irregularity/thickening developed in the pulmonary (n=1), aorta (n=1), mesenteric (n=1) and renal (n=1) arteries. 1 patient in this age group underwent initial imaging and was found to have renal artery stenosis and no calcifications.5 patients over 3 years of age underwent imaging (3 follow up and 2 initial), none of which had calcifications. Both initial studies in older siblings were normal. The follow up studies showed: stable infrarenal aortic, mesenteric and renal artery stenosis (n=1); progressive renal artery stenosis (n=1); and improved renal artery stenosis (n=1). Purpose or Case Report: With newer generation scanners providing free-breathing, high-resolution cardiac imaging at sub milli-Sievert radiation dose, demand for pediatric cardiac computed tomography angiography (CTA) is increasing. It is only a matter of time before sub-specialized emergent cardiac CTA service during weekends and after-hours on weekdays will become an expectation for optimal patient care. The purpose of this study is to describe our experience of providing this service and its effect on patient care.

We retrospectively identified all patients that underwent after-hours weekday or weekend emergent cardiac CTA between January 2017 and August 2018. Sub-specialized cardiac CTA in the setting of congenital heart disease and coronary imaging were included. Routine vascular imaging in patients without structural heart disease, such as for aortic dissection, is commonly handled by non-cardiac trained radiologists and was therefore excluded. Data collected included day and time of CTA, patient age, indication, patient location, post-op status, need for anesthesia, need for surgery, intervention and/or change in medical management based on the emergent CTA. Results: A total of 47 studies were identified, 26 (55%) of which were performed on a weekend or holiday and 21(45%) after 5 PM on a weekday. These were performed under direct supervision of a cardiac imager (1 of 3 FTE's). 28 (60%) were male. 34 (72%) were in infants out of which 19 (40%) were neonates. 16 (34%) patients were imaged for coronary artery evaluation, 9 (19%) for post-op conduit and shunt evaluation and 8 (17%) for aortic arch. Only 7 (15%) patients needed sedation with breath-holding for CTA, all of which were infants needing evaluation of coronaries. One patient was already intubated prior to CTA. 27 (57%) patients were from the ICU, 9 (19%) from intermediate care unit and 11 (23%) from the emergency room. Half (n=24, 51%) had at least 1 surgery for congenital heart disease at time of imaging. 29 (62%) CTA's had positive findings explaining patient's symptoms. Following CTA, 20 (43%) patients underwent either surgery or an interventional procedure, and 9 (19%) had change in medical management based on CTA findings. Surgery/intervention delay from CTA ranged from 0 days to 29 days with median of 3.5 days.

Conclusions: Emergent pediatric cardiac CTA is a valuable service leading to a change in management in 62% of our cases. Larger multi-institutional studies will be needed to ascertain if outcomes are affected by this service. Purpose or Case Report: Optimal coronary artery opacification in pediatric coronary artery CTA (CCTA) is highly dependent on appropriate administration of contrast material. There is no standard prescription for contrast volume, injection duration (ID), injection rate (IR), and timing in pediatric CCTA. We sought to assess the efficacy of an intravenous contrast administration protocol based on a fixed ID and either a weight-based IR (< 50kg) or fixed IR (>50kg) to achieve optimal coronary enhancement in pediatric CCTA.

Patients who underwent CCTA at our institution from 2015 to 2018 were retrospectively evaluated. Studies with a fixed ID (14 or 16 sec) and a weight-based or fixed IR were included. Timing bolus was utilized in all studies. Age, weight, indication, contrast administration parameters (ID, IR, contrast volume, additional delay from timing bolus peak enhancement), and acquisition parameters (kV, scan time, CTDI volume, and DLP) were recorded. Vascular enhancement was measured as mean attenuation in Hounsfield units (HU) at the aortic root, proximal right and left coronary arteries, distal right coronary artery (RCA), and descending aorta. Optimal enhancement was defined as > 350 HU. The presence and location of artifact from the incoming contrast bolus was noted. Results: 109 CCTAs were performed in 108 patients. Median age was 13.5 years (range 4 -20). Median weight was 51.5 kg (range 20.5 -118). Injection duration was 14 sec in 50% and 16 sec in 50%. Median injection rate for patients <50 kg was 0.1 mL/kg/sec. For patients >50 kg, the injection rate was 5 mL/sec in 58% and 5.5 mL/sec in 35%. Median contrast volume (including the timing bolus) was 1.9 mL/kg (range 1.0 -2.6). No patient had significant bolus artifact in the superior vena cava or right heart obscuring the coronary arteries. In 5%, there was mild or moderate bolus artifact at the level of the right pulmonary artery. Optimal enhancement in the aorta and proximal coronary arteries was achieved in 95% of the patients. Suboptimal enhancement of the distal RCA was seen in 11% of patients with optimal opacification of the proximal coronary arteries. Enhancement was higher in patients scanned at lower kV. Conclusions: Optimal coronary artery opacification can be achieved in pediatric patients utilizing a fixed injection rate and an injection duration protocol of 14 or 16 seconds. The mean contrast dose was 59.3 ± 34.5 mL. Iohexol was the most commonly used contrast medium (98.6%) followed by Iodixanol (1.3%). The mean peak pressure was 68.9 ± 62.3 psi and the flow rate was 1.7 ± 0.9 mL/s. Regarding patients with extravasation, the mean age was 11.2 ± 6.2 years; the most common peripheral IV access site was the antecubital fossa (n = 12). The most common catheter size was 22 gauge (n = 8) and the median catheter dwelling time was between 0 -3 days. The contrast agents used were Iohexol 350 mg/dL (n = 12) and Iohexol 300 mg/dL (n = 6). The PIVIE score observed were seven mild, six moderate and five severe. There is no difference in PIVIE score with respect to flow rate, peak pressure, catheter dwelling time, catheter size, catheter injection site, ultrasound guidance and contrast agent (p > 0.05). Cases with extravasation had significantly higher peak pressure (p < 0.001) and flow rate (p < 0.001) compared to those without extravasation. Patients who received Iohexol 350 had a significantly higher incidence of contrast extravasation compared to those who used Iohexol 300 (p = 0.03).

The use of power injectors in children undergoing contrast-enhanced CT is safe under current standards. However, a significantly higher rate of extravasation was seen in patients receiving contrast with a higher viscosity. Purpose or Case Report: Patent ductus arteriosus (PDA) stent placement in infants with ductal dependent pulmonary blood flow is being increasingly adopted in clinical practice. PDA stenting can be challenging due to variable anatomy. The purpose of this study is to correlate computed tomographic angiographic (CTA) morphology and length of the PDA with catheter angiography (CA) and its relation to eventual PDA stent length.

We retrospectively identified all pediatric patients that underwent PDA stenting at our institute from 2004-2017. Patients who had CTA prior to stenting were included. PDA length was measured by a radiologist blinded to the CA data, using Syngo-via post-processing software (Siemens, Germany). Vessel centerline technique was used. The actual length of the duct as well as straight length between aortic and pulmonary ends was measured. The PDA morphology was classified as straight (type I), mildly tortuous with 1 turn (II) and tortuous with >1 turn (III). The PDA was also measured and morphology classified on catheter angiography by an interventional cardiologist blinded to the CTA findings. The CTA and CA lengths, straight lengths and stent length were compared using scatter plots and Pearson correlation. Purpose or Case Report: The aim was to assess the blood flow distribution in the fenestrated and completed Fontan statuses as compared to normal using phase contrast magnetic resonance and to correlate them with abdominal MR imaging findings.

We identified 39 children with fenestrated or completed Fontan circulation in whom phasecontrast velocity mapping of major thoracic and abdominal vessels was a part of the magnetic resonance (MR). The patients were divided into 3 groups: fenestrated Fontan group with MR under general anesthesia (GA) (15 patients, average age 5.87 years), completed Fontan group with MR under GA (6 patients, average age 8.74 years) and completed Fontan group with MR without GA (18 patients, average age 12.52 years). Patients' flow data were compared with the previously published flow data obtained in heathy control (average age, 13.58 years) at fasting status and after a meal. Results: As compared to healthy controls, Group 1 (fenestrated Fontan) showed normal or marginally increased cardiac output (3.92 ± 0.40 vs 3.72 ± 0.69 L/min/m 2 , p<0.30), while Groups 3 (completed Fontan) showed decreased cardiac output (3.24 ± 0.71 vs 3.96 ± 0.64 L/min/m 2 , p=0.003). Both Group 1 and Group 3 showed reduced abdominal blood flow. Reduced abdominal flow was mainly due to diversion of a large part of cardiac output to aortopulmonary collaterals in Group 1 and reduced cardiac output in Group 3. Superior mesenteric and portal venous flows were more profoundly reduced in Group 3 than in Group 1. On the contrary, the hepatic arterial flow was mildly increased in Group 1 and markedly increased in Group 3. Group 2 showed flow data similar to Group 1 except for the tendency toward decreased cardiac output and superior vena caval flow. Hepatic parenchymal changes were more severe in complete Fontan than in fenestrated Fontan group. There was no significant correlation between the hepatic flow and imaging findings. Conclusions: Fenestrated and completed Fontan circulations showed distinctly different pattern of flow distribution among body parts. Further prospective study in a larger cohort is required to correlate the flow redistribution in Fontan physiology with imaging and laboratory findings as well as the long term outcome. There was a significant decrease in lymphatic malformation volume (p=0.0001), with a mean decrease of 39±25%. Four of 23 patients (17%) saw an apparent increase in lesion size, however the volume increase was relatively small (median 12%). When normalized to body mass index, 20/23 patients (87%) showed a mean 44% reduction in lesion volume. There was no correlation between percent volume decrease and age, sex or dose duration. Lesions demonstrated a change in tissue composition (lower relative fluid content) in 14 patients (64%), while unchanged in 8 patients (36%). Notably, 75% of increasing lesions (3/4) also had a change in tissue composition, whereas patients with volume reduction had a composition change in only 39% (7/18, p<0.005). Conclusions: Sirolimus therapy produced a significant reduction in lesion volume and relative fluid content in a majority of patients, suggesting efficacy in treatment of complex lymphatic malformations. Our results support consideration of sirolimus in the treatment of lymphatic malformations, but additional data that better predict responders, durability of effect, and long-term adverse effects would be beneficial in further characterizing its efficacy. Purpose or Case Report: To assess the feasibility of performing multiple complex vascular interventions on an anatomically accurate 3D-printed pediatric endovascular phantom.

The phantom consists of three anatomic sections, chest, neck, and thigh with arterial and venous vasculature created from CT angiography images of 15-year-old patients. The vasculature was aligned with chest, neck and thigh regions segmented from CT images of a teenager in Materialise Mimics and 3-Matic. Rubber tube connections inserted between the anatomic regions together with pumps simulate circulatory flow. Vascular access points were created at the neck and thigh region, covered with a replaceable "skin patch" for reusability. The vasculature mold and breakaway cavities were 3D printed on a Stratasys Fortus 450mc in ABS plastic. Smooth-On Ecoflex 30 silicone was used for molding simulating body texture and DragonSkin10 used for the skin patches.Three pediatric interventional radiologists and a trainee successfully performed the following procedures under fluoroscopic guidance: renal angiography, cavography, femoral and jugular supra and infra-renal IVC filter placement and retrieval, nontarget coil embolization and snare retrieval, and iliac stent placement. Post simulation, all operators completed a questionnaire, consisting of 13 questions, rating their experience with the phantom using a 5-point Likert scale (1-Strongly disagree, 2-Disagree, 3-Neutral, 4-Agree and 5-Strongly agree). Results: All operators reported that the 3D printed phantom and vasculature were anatomically accurate, and sheath placement, catheter/wire manipulation, supra/infrarenal filter placement and retrieval, and iliac stent placement were realistic. Three operators reported that the aortography and cavography compared favorably to live patients. All operators recommended the 3D printed phantom for basic vascular interventional training for fellows and for training in complex/infrequent pediatric vascular interventions for attendings. Conclusions: To our knowledge this is the first pediatric endovascular multiuse phantom providing capability of simulating complex/uncommon procedures in one model with "soft tissue" consistency. The next iteration will involve prospective enrollment of IR trainees and formal assessment of technical skill development with repeatability possible by replacing anatomic phantom segments (stenting). Initial feedback from operators validates the importance of the utility of the 3D printed medical model for training purposes. Results: Nine-patients (mean 13.9 years-old; range 6-19) with massive/submassive PE who underwent PA-CDT were included. PE was diagnosed by CT-angiography and CDT was technically successful in all cases. At cessation of CDT, followup pulmonary-angiography revealed complete thrombus resolution in 4-patients, partial resolution in 5-patients. Mean PA pressures decreased in all patients (mean pre-CDT PA pressure=36.5 mmHg; mean post-CDT PA pressure=28.0 mmHg). CDT alone was clinically successful in 7-patients (78%). One-patient with acute-on-chronic PE with severe pulmonary hypertension required surgical thrombectomy of chronic-thrombus after CDT of acute-thrombus. One-patient died following cessation of CDT for reasons unrelated to CDT procedure. There were no immediate bleeding complications from CDT therapy. All patients who survived were maintained on anticoagulation treatment following CDT. Conclusions: PA-CDT is a technically feasible and relatively safe therapeutic option for children and adolescents with submassive and massive pulmonary emboli. . NT-MTS clinical presentation ranged from lower extremity swelling, pain, and/or venous congestion. All patients underwent pre-and-post stent intravascular ultrasound (IVUS) and venography. All T-MTS had stent placement performed at the time of pharmacomechanical rheolysis. Catheter directed thrombolysis with pharmacomechanical rheolysis was performed in all cases of TMTS. All patients were treated with self-expanding stents, ranging from 14-22mm in diameter, and 4-9cm in length. Patients with residual thrombus after pharmacomechanical thrombolysis were placed on our institution's catheter directed thrombolysis protocol using weight-based dosing of tPA; repeat venography and IVUS were performed the following day to evaluate for residual clot burden. Clinical follow up was conducted at 1-6 month time points. Patients with residual or chronic thrombus or persistent/recurrent symptoms received follow-up lower extremity Doppler sonography. All patients with T-MTS were treated with at least 6 months of anticoagulation by the hematology service. All patients with NT-MTS were treated with 6 months of anti-platelet therapy. Results: Technical success was 100%. Clinical success (defined as improvement/elimination of symptoms) = 90%. Overall patency at 6, 12, 18, and 24 months was 100%. No patients had recurrence of symptoms. One periprocedural complication included histaminergic reaction to paparavine given for venospasm.

Conclusions: Endovascular stenting of the common iliac vein in thrombotic and nonthrombotic MTS is technically feasible and clinically efficacious with excellent patency and low complication rates in adolescent patients. Continued retrospective analysis of patients in our database is ongoing. Results: Technically successful PV recanalization and reduction of PV pressure gradient to ≤ 5 mmHg was performed in 17/18 patients (94%). Technical success in split liver vs whole liver was 64% (9/14) and 86% (12/14) in the 28 attempted and successful recanalization procedures respectively. PV Recanalization success was higher at first attempt in the whole liver group (10/12, 83%) in comparison to the split liver group (3/6, 50%). There was improvement in signs and symptoms of portal hypertension in 16/18 (89%) patients. Both procedure time (median: 4.3, range: 2.3-9.9 hours) and fluoroscopy time (median: 36.9, range: 7.4-153 minutes) were markedly greater in the split liver group, while the whole liver group was a median of 3.1 hours (range: 1.4-6.8) and 24.8 minutes (range: 5.2-113.9) respectively. . The procedure time (median: 5.9, range: 4.3-9.9 hours) and fluoroscopy time (median: 66.1, range: 38.6-153 minutes) were also both higher in the split liver group during the time of successful recanalization in comparison to the whole liver procedure (median:3.5, range: 1.8-6.1 hours) and fluoroscopy time (median:46.9, range: 8.4-75.5 minutes). Age (median: 3, range: 2-16 years) and weight (median: 17, range: 13-47 kg) are both reported lower in the split liver group in comparison to whole liver with median age of 5 years (range: 1-17) and weight of 28 kg (range: 21-71) respectively.There were 7 (17%) major complications (SIRC) total. In split liver vs whole liver there were 4 (20%) and 3 (22%) complications respectively. These included perisplenic hematoma (n=2), hemoperitoneum (n=4), and hepatic artery pseudo aneurysm (n=1) managed with pain medication and blood product replacement. Median follow-up was 17 months (range: 2-84 months). Conclusions: Technical challenges requiring longer procedure and fluoroscopy times could be contingent upon the anatomy of the patient based on the type of OLT. Though small sample, technical and clinical success were comparable. were underwent tunneled C-line or PICC insertion. Both techniques were performed by a single pediatric radiologist. The procedure time and catheter lifespan between the two techniques were compared using the Mann-Whitney test. The rate of bedside procedure and catheter-related bloodstream infection (CRBSI) between the two techniques was calculated. Results: 24 cases (mean age 5 years, range 0-15) of tunneled Cline insertion and 8 (means age 6 years, range 2-18) cases of PICC insertion were performed. The procedure time was significantly shorter in tunneled C-line insertion than PICC insertion (13±5 min Vs. 19±6 min, p=0.030). The catheter lifespan showed no significant difference between tunneled Cline and PICC (12±11 days Vs. 24±21 days, p=0.192). The bedside procedure was performed for poor patient's condition in 88% and 0% of cases of tunneled C-line and PICC, respectively. The rate of CRBSI of tunneled C-line and PICC were 6.7 and 10.6 per 1,000 catheter-days, respectively. Conclusions: Tunneled C-line was associated with shorter procedure time, no significant difference in catheter lifespan and a lower rate of CRBSI compared with PICC. Tunneled C-line can be an alternative to PICC in severely ill young children at the bedside. Purpose or Case Report: The purpose of our study was to test our custom software program's ability to estimate the blood flow velocity envelope for transcranial spectral Doppler ultrasound waveforms.

Our group has programmed a custom algorithm that uses image contrast and neighboring data to estimate the envelope of a spectral Doppler waveform. Our algorithm was implemented in java and uses image-based extraction methods: region of interest reductions, preprocessing, and a novel envelope extraction technique. To test this method against industry standards, we performed a retrospective review on pediatric patients who underwent transcranial Doppler between 1/1/2008 to 7/15/2017 at a single pediatric center on Philips IU-22 machines. Ultrasound images of these waveforms formed three groups: waveforms with manufacturer envelopes, waveforms with technologist drawn envelopes and waveforms without envelopes. The manufacturer and manual envelopes were removed from those images and a human drew the ideal waveform envelope on each waveform. The average absolute pixel difference between the human gold standard and another technique (custom software, manufacturer, and technologist) was the metric of comparison and was also scaled to the region's size. The entire waveform and a single manufacturer determined wave region were compared across techniques. Correlation analysis was also performed on the average pixel error between our algorithm and the manufacturer algorithm for each waveform. Results: 230 patients were included with 3810 unique waveforms. Our approach outperformed the manufacturer or technologist by average pixel error compared to the gold standard (p<.01). For entire waveforms, our algorithm was on average 2 pixels more accurate than the manufacturer; for single waves, our algorithm was on average 2.9 pixels more accurate than the manufacturer and 1 pixel more accurate than the technologist. For entire waveforms unable to be processed by the manufacturer, our algorithm was on average 18 pixels different from the gold standard. Error correlation was low between our approach and manufacturer (r 2 <.32). Artifacts and resizing seemed to have little effect on our technique, with waveform quality impacting measurement more than artifacts or scaling. Conclusions: Our custom algorithm outperforms manufacturer techniques and technologists' manual tracing in determining accurate waveform envelopes for TCD images. This work is a precursor to using these waveforms for input into machine learning algorithms. Conclusions: Type 2 annular fissures are associated with cervical cord ASA infarct in children and are uncommon in children without ASA infarct (2-5%). Disc dessication is somewhat more common in children with ASA infarct but is commonly seen in children without ASA infarct (41-54%). Interestingly, annular fissures associated with ASA infarct typically do not enhance at presentation but may develop enhancement on follow up. Non-enhancing type 2 annular fissures may be acute and are a potential pathway for fibrocartilaginous disc emboli to reach the ASA and cord. Two studies were performed for thoracic outlet syndrome. In multiple cases, findings on MRI with ferumoxytol were confirmed with biopsy and conventional angiogram. None of the patients had an adverse reaction to the ferumoxytol. Conclusions: Ferumoxytol-enhanced MRA is a commonly used contrast agent at our institution to diagnose vascular abnormalities in the brain or whole body. Use of ferumoxytol allows the radiologist to interrogate multiple territories in one study due to its highly stable intravascular time. We were able to confidently detect and characterize vascular abnormalities and exclude vascular malformations. No infusion reactions were noted in our cohort. Therefore, ferumoxytol is a beneficial and safe alternative to gadolinium-based contrast agents for high resolution CNS and whole body MR angiography. 3 minutes) and a highly accelerated Wave-CAIPI SWI sequence (R=9, TA=1.0 minute for the 32-channel head coil; R=6, TA=1.5 minutes for the 20channel head coil). Two blinded neuroradiologists independently rated both sequences in terms of artifacts, noise (central and peripheral), anatomic contrast (ability to delineate normal structures) and pathologic contrast (ability to delineate the clinical pathology, when present). The images (standard SWI and Wave-CAIPI SWI) were anonymized and presented side-by-side on two adjacent monitors. A predefined 5-points scale was used, where negative numbers favor standard SWI and positive numbers favor Wave-CAIPI SWI: 0 indicates that the sequences were equivalent, -1 or 1 indicates that one sequence was subjectively preferred but the difference would not have changed the diagnosis, and -2 or 2 indicates that one sequence was superior and the difference would impact the diagnosis. Wilcoxon signed-rank test was used to evaluate the difference between sequences for each variable.

Results: There were no failed exams (e.g., due to excessive motion or inability to tolerate scanning). The most common indications for MRI included tumor, epilepsy, and headache. Standard and Wave-CAIPI SWI had similar performance for evaluation of pathologic findings, normal anatomic structures, noise within the peripheral brain, and artifacts. There was subjectively more noise within the central brain (n = 15, P <0.001) using the Wave-SWI sequence, however there were no cases where this difference would have impacted the final diagnosis (no scores of +/-2). The inter-reader agreement for these parameters was moderate to good (kappa 0.45 -0.75).

Highly accelerated Wave-CAIPI SWI of the brain is feasible in non-sedated children and can be achieved without significant impact on overall diagnostic quality. The accelerated sequence was associated with a subjective increase in central (but not peripheral) image noise. Results: First pass contrast enhanced MRV with subtraction provided the best visualization of cortical and briding veins, even in infants as young as 2-3 months. 3D phase contrast MRV was nearly as good, but image quality was scanner dependent. Sagittal 2D TOF MRV was moderately good at demonstrating cortical and bridging veins near the sagittal sinus, but suffered from in-plane flow saturation and poor spatial resolution in coronal or axial reconstructions. Coronal 2D TOF MRV was the least desireable for briding vein evaluation. Weighted MIP images provided the best (and easiest) overall visualization of the superficial venous system. Thin-slab MIP and volumerendered techniques were useful in selected cases. Conclusions: Selection and optimization of MR venography pulse sequences and post-processing techniques allows for the reliable demonstration of cortical and bridging veins in infants and young children with suspected abusive head trauma. Improved characterization of bridging vein injury may distinguish traumatic from atraumatic SDH and provide valuable insight into the nature and magnitude of head injury. Purpose or Case Report: PHACE syndrome is defined as a large segmental infantile hemangioma (IH) larger than 5 cm, posterior fossa malformations, cerebral arterial anomalies, aortic coarctation, eye anomalies and sternal defect. Cerebrovascular anomalies are estimated at 55% and cardiovascular anomalies at 42%. Cerebral artery anomalies were reported to be on the same side as the facial IH. The objective of our study is to review facial hemangioma and the risk of cerebral or cardiovascular anomalies in PHACE or non-PHACE patients.

We reviewed all patients seen in our vascular anomalies group with the diagnosis of facial IH and who have clinical photos and MRI. The clinical photos were reviewed to determine the localization of the lesion by segment and the right or left lateralization. Cerebrovascular, posterior fossa or cerebral anomalies were reviewed. The presence of cardiovascular anomalies, sternal defect, eye anomalies or intracranial hemangioma was recorded. Two experienced paediatric radiologists scored brain MRI. Clinical data was reviewed by dermatologists using the consensus statement on diagnostic criteria for PHACE syndrome. Results: One hundred twenty-six children (93 girls and 33 boys, mean age 16.8 months) with facial hemangioma and brain MRI scan were included. Twenty-one patients out of 126 had PHACE syndrome. Thirty-eight (30%) children had a segmental hemangioma larger than 5 cm and 88 (70%) smaller than 5 cm. Cerebrovascular anomalies were seen in 18/126 (14%). Of them, 17/18 (94%) were PHACE. Brain anomalies were seen in 9/21 PHACE patients (42%) whereas only one non-PHACE patient had brain anomalies (p<0.001). Seventeen of 38 patients with hemangioma larger than 5 cm had an associated cerebrovascular anomaly whereas only one patient with a hemangioma smaller than 5 cm had cerebrovascular anomaly (p<0.001). Eight of those 38 patients (30%) had associated brain anomalies whereas only 2 patients with hemangioma of less than 5 cm had associated brain anomalies (p<0.001). The laterality of the cutaneous hemangioma and the underlying cerebrovascular anomalies was concordant in 61% (11/18). Cardiovascular anomalies were seen in 6 patients and ocular anomalies in 8 patients. All of them had a PHACE syndrome.

In patient with facial segmental hemangioma of more than 5 cm, brain MR, MRA and cardiovascular MR have to be performed. The ipsilaterality of the cerebrovascular anomalies was seen in only 61% of our cases. Long-term follow-up is recommended. Purpose or Case Report: Bone Age (BA) is the measurement of the skeletal maturity of an individual using hand radiograph. Manual BA assessment, performed by a trained radiologist, is the current gold standard and has high inter observer variability (IOV). BoneXpert is an automated machine learning algorithm (software) with zero IOV, and generates results in 5 seconds. The aims of this study are to compare the BA analysis by BoneXpert with the radiologists' manual readings in a crosssectional cohort, as well as a longitudinal cohort. manual readings.

A total of 614 BA studies from 473 patients, assessed by three radiologists from 2013-2018 were identified for automatic determination of BA by BoneXpert (Visiana, Holte, Denmark). BA readings by BoneXpert were compared with those by radiologists' manual readings based on the Greulich and Pyle Method. Spearman's correlation coefficients were used to assess the association between BoneXpert and manual readings. Results: A total of 583 BA studies were included in the crosssectional analysis comparing the manual readings and the BoneXpert's assessment. BA data from a subset of 119 patients that had two or more BA studies were analyzed for a longitudinal study to compare the change in skeletal maturation between these 2 methods. BA analyses by BoneXpert and manual reports showed a strong correlation (r = .9786; P < 0.0001) with the mean difference of BA between these 2 methods being 0.12 ± 0.76 years. When stratified to male and females groups, BA analyses by BoneXpert and manual method remained strongly correlated (P<0.0001) in each gender group.

In the longitudinal study, there was also a strong correlation between BoneXpert and manual readings (r = .7852; P < 0.01). Conclusions: BoneXpert can analyze the hand radiographs quickly and effectively, and produce a BA assessment that is comparable to the manual readings. Other future research will determine the clinical utility of Bone Health Index assessed by radiogrammetry of BoneXpert and how it is correlated with bone density measurement and bone health in pediatric patients with various metabolic bone disorders.

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Skeletal bone age assessment is a common clinical practice to investigate endocrinology, genetic and growth disorders of children. Clinical interpretation and bone age analyses are time-consuming, labor intensive and often subject to inter-observer variability. Bone age prediction models developed with deep learning methodologies can be leveraged to automate bone age interpretation and reporting. The bone age model developed at our institution was offered to interested health systems and institutions to implement and validate the model. This study discusses the logistical, technical, and clinical issues encountered with this model implementation.

After IRB approval, multiple U.S. based radiology departments were solicited to adopt and validate the Stanford University bone age model. A total of 8 institutions (4 standalone pediatric hospitals and 4 academic radiology departments) agreed to partner with the primary investigators. IRBs at each institution were required in addition to registration with ClinicalTrials.gov registry. Standardization of the data use agreements was performed. Patient data and protected health information data was retained at each institution. Technical requirements included model hosting at each institution and integration to send images to the model server and results to the interpreting radiologists. Results: Multiple logistical, technical, and clinical issues were encountered. IRBs at the various institutions had different requirements including waiving patient consent. Technical differences between institutions included model hosting, PACS integrations, interfaces with the reporting system, and image preprocessing. Clinical differences included report templates, calculation of bone age standard deviation, use of Brush foundation, and ability to directly send bone predictions to the reporting system (versus displaying the results as a separate interface). The bone age model was successfully implemented at 7 institutions and approximately 190 studies have been evaluated.

Conclusions: There are myriad challenges to implementing and validating models developed with deep learning methodologies. As models are developed for various clinical use cases including bone age assessment, it will be incumbent on radiology practices and health information systems to integrate these models into clinical practice. Conclusions: Our feasibility study on the use of CNN for CAD of tibial fractures has the promise to automate and improve workflow for radiologists. Our partly-trained CNN network using a limited dataset, similar to the numbers a typical radiology trainee will see, was robust and able to multiorthogonally identify fractures and distinguish these fractures from normal physes. This pilot exam on tibial fractures will serve as a basis and foundation for future CNN training for long bone fractures elsewhere in the skeletally immature child. Purpose or Case Report: Radial head and neck fractures represent a small fraction of all pediatric elbow fractures but are often associated with significant long-term morbidity. Various mechanisms can cause these radius fractures; however, the constellation of associated injuries is poorly understood in the pediatric population. The goal of this study is to assess the frequency of additional fractures in the setting of proximal radial head and neck fractures.

This IRB-approved study reviewed 600 consecutive cases with reports suggesting a proximal radius fracture obtained during a five-year period (2013-2017). Pathologic and non-radial head or neck fractures were excluded. Initial and follow-up radiographs were re-reviewed to assess the presence and type of fracture as well as the presence of elbow joint effusion on initial presentation. Fracture types and properties were analyzed using ANOVA and independent samples t-tests. Results: 334 radial head and neck fractures were identified (50% female; mean age: 9.6 years, SD 3.9). 37% were noted to have one or more additional fracture about the elbow. Olecranon fractures were most common (23% of all cases), followed by additional medial epicondyle fractures (4%), coronoid fractures (3%), ulnar diaphyseal fractures (3%), and olecranon with medial epicondyle fractures (2%) and others (1-2%). Approximately one-third (33%) of additional fractures were missed by the initial interpreting radiologist. Most (68%) of the missed fractures were visible on the initial radiographs, but 32% were only visible on follow-up radiographs because of Digital tomosynthesis (DT) is an imaging technique that creates multiple radiographs with high in-plane resolution. The hypothesis of this study is that DT can assess displacement of medial epicondyle fractures with similar accuracy to CT and lower radiation dose. Radiopaque 3D-printed models were used to optimize positioning and DT technique.

Three 3D-printed models were generated based on CT data of pediatric patients with medial epicondyle fractures using at least 20% polylactic acid fill. DT was performed in thirteen projections/sweeps to determine optimal visualization of fracture fragment displacement. Two pediatric radiologists independently reviewed all projections/sweeps and measured maximal displacement. CT of the 3D-printed models was performed as the gold standard with consensus agreement by the two radiologists on maximal displacement. Estimated radiation dose for pediatric elbow DT was calculated using phantoms. Non-contrast upper extremity CT dose estimates were extracted from dose reports at the authors' institution between July 1, 2015 and June 30, 2018 (N=52). Results: Absorbed dose estimates for DT were 0.15-0.21 mGy, compared to 0.02-0.04 mGy for 2-view elbow radiographs and median CT upper extremity dose 3.9 mGy. Six projections were excluded for poor fracture visualization and inter-reader measurement variation >1 mm. The remaining seven projections had good inter-reader agreement (both readers <1 mm variance from each other, Pearson coefficient 0.93). The lateral transverse and oblique longitudinal DT projections were closest to consensus CT maximal displacement (2.9 and 3.4% measurement variance from CT, respectively). Conclusions: 3D-printed models help to optimize DT technique for ideal visualization of displacement of pediatric medial epicondyle fractures. The lateral transverse and oblique longitudinal DT projections had high inter-reader agreement and high correlation with CT measurement of fracture displacement, at approximately 1/20 th the absorbed radiation dose. Purpose or Case Report: Hutchinson Gilford progeria syndrome (HGPS or Progeria) is a rare sporadic genetic disorder (1 in 20 million) that uniformly is fatal, with most deaths attributed to myocardial infarction between 7-21 years of age. Due to the disease rarity and short patient life-span, there are only ~400 children with Progeria alive worldwide at any given time. One early clinical manifestation in Progeria is abnormal skeletal growth, yet this has not been fully characterized. The objective of this study is to characterize the longitudinal long bone growth and skeletal maturation pattern of children with Progeria.

Skeletal survey radiographs of children with Progeria performed at a large tertiary hospital (1/1/2009-7/31/2018) were reviewed. The protocol for the surveys varied over time, but the majority included PA radiographs of the left hand/wrist, and AP radiographs of each humerus, radius, ulna, tibia, and fibula. No femur radiographs were obtained. Using the left hand/wrist radiographs, bone ages of these children were estimated by the standards of Greulich and Pyle. Following methodology established in the literature for studying long bone growth, the study cohort was separated into 2 overlapping age groups: childhood (≤12 years-old) and adolescence (≥10 yearsold). For those in the childhood cohort, bone length measurements were made between physes along the midline long axis of the bone. For those in the adolescence cohort, bone length measurements were made along the midline long axis of the bone from the upper margins of the proximal to the lower margin of the distal ossified epiphyses. These long bone length measurements (humerus, radius, ulna, tibia, and fibula) and bone age estimates were plotted against patient chronologic ages and compared to published normal reference standards. Results: Eighty-six children (M:F=46:40) with 253 skeletal surveys (M:F=121:132; age range=2 months-22 years) were included. Bone age estimates showed a skeletal maturation pattern that is comparable to normal. However, longitudinal long bone lengths from these children began to significantly deviate from normal by age 1-2. The growth curves of these long bones plateau at around age 10. At adulthood, long bone lengths ranged from 48-61% of normal. Conclusions: Our study establishes reference standards for long bone growth and skeletal maturation of children with Progeria. These 2 outcome measures may offer objective assessments to evaluate the efficacy and response of these children undergoing treatment.

Can Ultrasound be reliably used to evaluate infants with DDH after age 6 months without the use of plain film radiography?

Noor M. Maza Purpose or Case Report: DDH occurs in 1.5/1000 births and requires prompt diagnosis and treatment as well as interval clinical and imaging surveillance for assessment of healing and complications. Current screening and monitoring guidelines recommend the use of US for evaluation of the hip in infants <6 months of age, and XR imaging in patients >6 months. The purpose of this study is to determine whether US is a reliable surrogate for XR for evaluation for DDH in patients >6 months.

This is an IRB approved retrospective study of a series of 22 at risk patients >6 months of age who had both US and XR for assessment of DDH. All of the imaging studies of potential patients were reviewed by 1-2 experienced pediatric radiologists with all the studies having been performed by an experienced pediatric radiologist. 30 patients >6 months were identified who had both US and XR studies of which 8 patients with a time difference between modalities greater than 4 months were excluded. Dynamic hip US was performed (transverse flexion and longitudinal flexion views without stress and with stress and during the Ortolani maneuver were obtained; Graf angles measured), except when patient was in a Pavlik harness or brace at which time adynamic hip sonography was performed. When it was determined that alignment was normal without stress, dynamic hip US was added. Size, configuration, and symmetry of the femoral capital epiphyseal ossification centers was assessed when present. All of the studies were performed using Philips IU22 US units (L12-5 and/or L9-3 probes with a few of the earlier images obtained with a C5-2 or 9-4 probe). The US reports included descriptions of the hip morphology (including Graf angles), alignment, and stability. AP supine XR included either a neutral projection or both neutral and frog lateral views. The shape of the acetabulum, alignment of the hips, and the presence, size and symmetry of the femoral capital epiphyses was assessed, and the acetabular angles were measured on neutral views. Alignment was also assessed on frog lateral views. Results: In all 22 patients, there was excellent correlation between the US and XR finding. Conclusions: US may be used as a first line imaging modality in babies >6 months of age presenting with clinical suspicion of DDH as well as those who were diagnosed with DDH prior to 6 months and are due for follow up surveillance, as long as the operator is appropriately experienced and the probe technology allows for visualization of all pertinent landmarks.

Jeffrey Prince, MD,jeffreysprince@gmail.com; Department of Medical Imaging, Primary Children's Hospital, Salt Lake City, UT Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Care process models (CPM) give guidance for the use of imaging in clinical scenarios and can be effective in screening processes.A CPM needs to be supported by and modified based on outcomes measures. We evaluate 2 CPM's associated with screening for developmental dysplasia of the hip (DDH) in infants to see if the outcomes measures support or improve the CPM. One model gives guidance regarding timing of initial ultrasound evaluation for possible DDH in an attempt to avoid multiple ultrasound studies as part of the evaluation. The second model involves the methods used in evaluated patients with a breech risk factor for DDH.

All patients evaluated with ultrasound for possible DDH in a 27 month period were retrospectively reviewed. Age at the first ultrasound exam, result of that exam, number of follow-up exams, referral to orthopedics, and ultimate outcome for the patient were recorded. Additionally in patients evaluated because of breech presentation, the use of a follow-up radiograph between 6 and 12 months of age and the outcome of that study was also recorded. Results: Over the review period, 1830 patients were evaluated for the possibility of DDH. 143 were diagnosed with DDH. 1289 were felt to be normal. 398 were placed in an immature category. Immature categorization was given in 38.4% of patients under the age of 6 weeks at initial evaluation, in 29.4% of patients between the ages of 6 and 7 weeks, and in 7.3% of patients at 8 weeks of age or greater. Patients in the immature category required on average 0.92 additional ultrasound examinations. 97.9% of immature patients normalized without therapy.592 were evaluated initially because of breech presentation as a risk factor over a one year period. 203 of these patients had a follow-up radiograph as part of their evaluation as was recommended in the CPM. 98.0% of these patients were normal at the time of the radiograph. 12 patients were referred to orthopedics because of an abnormal radiograph. 4 of these required therapy and 8 were observed and remained normal clinically.

Outcomes measures support the recommendation that an initial hip ultrasound for risk factors should not be performed until the patient is 6 weeks of age. If this recommendation had been followed, 111 unnecessary ultrasound examinations could have been avoided. Outcomes measures do not support the recommendation of a follow-up radiograph for patients screened for a risk factor of breech presentation. These patients rarely develop hip dysplasia. Purpose or Case Report: We compare standard and low-dose modes of an X-Ray slot-scanner through a wide-ranging evaluation of dose and image quality for scoliosis exams.

The EOS TM slot-scanner is a bi-planar system that can acquire simultaneously two orthogonal x-ray images. The patient is scanned vertically by fan beams coupled with linear gaseous detectors. The system offers high quality images using radiation dose several times lower than DR due to a better management of scatter radiation.Recently, a new low dose mode was introduced for the system. The goal is to offer images of diagnostic quality using less radiation dose. This is achieved by lowering the tube tension and current while applying adapted image post-processing.In this work, we compare images acquired on 28 pediatric patients and their respective dose. Patients underwent two scoliosis follow-up examinations, both consisting in the acquisition of lateral and posteroanterior x-ray images, once with standard dose parameters and once in low-dose mode. Dose-Area product for each examination was collected and OSL dosimeters were placed near radiosensitive regions on the patient before each scan. Two radiologists evaluated the images in a randomized order using a questionnaire targeting anatomic landmarks. Visibility of the structures was rated on a 4 degrees scale. Image quality assessment was analyzed using a Wilcoxon signed ranks test.

Results: Dose-Area Products show decrease in dose ranging from 3 to 5 times when imaging in low-dose mode. These reductions are consistent with direct measurements obtained with OSL.All images meet diagnostic quality standards. Assessment of global image quality and visibility of a large majority of musculoskeletal landmarks showed no statistically significant bias when comparing standard and low-dose images.

Few landmarks, such as femoral heads in lateral view, were better seen in standard mode (p=0.013), while other, like cervicothoracic junction in posteroanterior view, scored better in low-dose mode (p=0.002).

We established that the low-dose mode offers diagnostic image quality while decreasing drastically the dose to the patient.This dose reduction is crucial for scoliosis followups since they involve repeated X-rays that can increase the risk of detrimental effects due to ionizing radiation. Purpose or Case Report: Flat panel detectors (FPD) have increasingly replaced image intensifiers (IIs) with the aim of providing high image quality and potentially lower patient radiation doses. As limiting radiation exposure is particularly important in our pediatric patients, our objective was to compare exposure rates and image quality between FPD and II fluoroscopes.

Exposure rates were measured for the FPD in both medium-and low-dose modes and for the II in lowdose mode at three magnification levels for both continuous and pulsed fluoroscopy at 7.5 frames/sec. All measurements were recorded at 1 cm above the fluoroscopy table. Image quality was evaluated using the 19.3 cm CIRS Rad Fluoro phantom, with the detector positioned 50 cm above the table on both the FPD and II to compare spatial and contrast resolution at the same 3 magnification levels.

Results: For pulsed fluoroscopy in low-dose mode, exposure rates (R/min) were lower for the FPD than the II at all mag levels:At 7.5 fps, mag factors 1, 2, 3, FPD values were: 0.79, 1.18 and 1.62 R/min.II values were: 1.36, 2.32 and 3.90 R/min.In medium-dose mode, measured similarly, FPD exposure rates were slightly higher than the II at the first mag level, but lower at the second and third levels: 1.44, 1.78 and 2.10.For continuous fluoroscopy at low-dose, exposure rates were lower in the FPD at all 3 mag levels: FPD 1.99, 3.11 and 3.8 vs II 2.78, 3.48 and 4.42.At medium-dose, FPD rates were lower than the II at the first mag level, but higher at the second and third levels: 2.55, 4.26 and 5.74.Spatial resolution between low-dose II to low-dose FPD was better on the II (4 objects vs 3 well-delineated). Between low-dose II to mediumdose FPD, resolution was better on FPD especially for mag 3 (4 objects vs 3 at mag 1; 5 vs 3 at mag 3).Contrast resolution between low-dose II to low-dose FPD was better on II (3.5-4 objects vs 2.5-3); between low dose II to medium-dose FPD, contrast was equivalent (3-4 objects for both).

Conclusions: Use of state-of-the-art fluoroscopy FPD equipment over the traditional II results in radiation dose savings when operating in low-dose mode, although this comes at the expense of some image quality. Changing to medium dose mode eradicates some of the radiation savings and results in modest gain in image quality which may not be commensurate.

Our findings indicate that care must be taken in initial equipment calibration, physicist input is important, and image quality should be tailored to the needs of the specific examination. Purpose or Case Report: Fluoroscopic exams are a significant contributor to pediatric patient radiation dose. Innovations such as "last image hold", pulsed fluoroscopy, and digital fluoroscopy have reduced the patient dose; however, development of new technology to further reduce dose has been largely absent in the last decade. A source of dose during fluoroscopy is "fluoro hunting" which is the use of fluoroscopy to center the image over the site of interest. The purpose of this study is to determine if the use of optical video to position the image before using fluoroscopy can reduce the dose from the exam.

An optical video device was designed and placed on the fluoroscopy unit used at El Paso Children's Hospital. The device was used on 20 consecutive voiding cystourethrograms (VCUG) and 20 consecutive upper GI (UGI) exams performed on children ages 2 weeks to 15 years by one pediatric radiologist. The device was used to center the fluoroscopy machine in relation to the site of interest prior to using radiation during these exams. The cumulative air-kerma area products (KAP) and fluoroscopy times (FT) were recorded for these exams. This data was compared to the same operator's KAP and FT from UGIs and VCUGs performed prior to the development of this study. A student's independent sample ttest was used to compare the KAP and FT from the exams using the device and those performed without the device. Results: FT was used as the primary indicator of effectiveness since KAP also depended on patient sizes. These FTs were compared to the retrospective control group. The average FT using the device for UGI was 3 seconds. The average FT from 115 UGIs without the device was 20.7 seconds. This was an 85.5% decrease in FT when using the device. The average FT using the device for VCUG was 4.5 seconds. The average FT from 78 VCUGs without the device was 22.8 seconds. This is an 80.0% decrease in FT when using the device. The student ttest revealed a statistically significant decrease in FT and KAP using the device.

Conclusions: This study revealed that using an optical video device to align patients for fluoroscopic imaging significantly lowers the FT of both UGI and VCUG. While at least a portion of the time reduction is likely due to the elimination of "fluoro hunting", the reduction was larger than expected. Since the device allows the use of fluoroscopy to be limited to actual diagnostic rather than alignment imaging, it also likely increases the radiologist's awareness of when radiation is being used. Simulated low-dose scans formatted in the axial and coronal planes as 3 mm reconstructions were presented to 3 observers (2 pediatric radiologists, 1 adult abdominal radiologist). The amount of added noise was selected based on prior responses of the individual (adaptive simulation). They marked the location of the appendix, rated the likelihood for appendicitis on a 6-point scale (1-definitely normal to 6-definitely appendicitis). The full-dose study was then presented to the observer, who again marked the appendix, rated the likelihood for appendicitis and the visualization of the appendix. The in situ noise in the appendiceal region was estimated using variance propagation methods. The intraobserver agreement of readers, matching between low dose and full dose, was fit to a parameterized psychometric function (agreement versus noise level) by maximizing the log-likelihood probability. This function can be used to analytically compute a noise shoulder Ν∑, which characterizes a reader's sensitivity to noise. Readers' responses for appendicitis and normal cases were also separately analyzed. Results: At full dose the observers had an area under the curve (AUC) of 0.973(0.018, stdev) and selected locations had a mean noise of 9.9 (4.3) HU. Individual readers had a range of noise tolerances (shoulder location), with mean tolerated noise of 23.6 HU (range 18.9-30.3). This equates to an 85% dose reduction. In general, appendicitis cases were less tolerant of noise reduction than normal cases.

The adaptive psychometric function is a promising tool to efficiently determine ALARA for taskspecific CT diagnosis, providing an order of magnitude reduction in cases needed compared with receiver operating characteristic curves. These results indicate that estimated CT dose decreases of 85% or a noise ceiling of 23.6 HU could be tolerated without affecting agreement. Purpose or Case Report: Abscess formation is a common appendicitis complication both pre-and post-operatively. At our institution, MR is exclusively used to evaluate for acute appendicitis after an equivocal or non-diagnostic US study, and contrast-enhanced CT is the test of choice only when abscess is suspected clinically or following US. Our aims are to 1) assess the radiation burden of patients who undergo CT for suspected appendicitis-related abscess and 2) determine the frequency of abscess detection and positive predictive value (PPV) for abscess detection of non-contrast appendicitis MR including diffusion weighted imaging in the setting of suspected appendicitis.

We performed an IRB approved retrospective review of all children who between 2015 and 2018 underwent CT after an US study suspicious for abscess in the setting of either suspected appendicitis or to evaluate for abscess post-appendectomy. CT effective dose was calculated using Monte Carlo simulation software (Radimetrics, Bayer Healthcare). We also reviewed all MR studies performed after an equivocal US for acute appendicitis to determine the frequency of abscess detection and the positive predictive value (PPV) of MR for abscess evaluation. Clinical course including surgery or interventional (IR) drainage was recorded for patients with abscesses by CT or MR.

Results: 220 CT studies in 165 patients (mean 9.3+/-3.8 years, 58% male) were performed for abscess evaluation with a mean CT effective dose of 9.8 mSv per exam or 13.1 mSv per patient. 35% of patients (57/165) had CT imaging related to initial appendicitis presentation of which 70% (40/57) were positive for abscess; 48% (19/40) went on to IR drainage, 45% (18/40) to surgery. 65% of patients (108/165) had CT imaging postappendectomy of which 66% (71/108) were positive for abscess; 76% (54/71) went on to IR drainage and 7% (5/70) to surgery. The patients who did not have IR drainage or surgery were treated conservatively with antibiotics. Of 893 MR exams for appendicitis, 20 were read as positive for abscess and all had proven abscess by operative or IR drainage, PPV of 100%. Additional ongoing MR data collection will be presented in detail.

Conclusions: There is a substantial CT radiation burden with an average of 13 mSv per patient to evaluate for abscess in the setting of appendicitis. MR has a high PPV for abscess detection suggesting it is a viable alternative to CT and allows for a radiation-free appendicitis imaging pathway. Results: Comparing all SSDL and SECT patients, there was CTDIvol dose parity (4.0 mGy vs. 3.6 mGy, p=0.0840), with individual weight subgroup dose increase of 5%-11% for CTDIvol and 2-12% for SSDE. SNR, CNR and IQ scores were maintained or improved in all SSDL weight groups with the exception of 15% reduced CNR in the 0-30 kg subgroup. We present representative cases which demonstrate how CNR can be recovered with low-energy virtual mono-energetic (mono-E) images (VMI) following material decomposition of the spectral data. Our measured SSDL doses were comparable or lower than those reported for children imaged using other dual-energy scanners and compared to the ACR national dose index registry for pediatric body CT, would rank in the lower quartiles of reported CTDIvol. Minor dose increase observed in patients despite matching AEC and dose index are likely secondary to machine differences in filters, collimation and a softer beam spectrum.

Conclusions: SSDL can be safely implemented in a pediatric department with overall dose neutrality and maintained image quality; however, there may be increased dose observed in the smallest patients despite matching AEC and dose index likely secondary to differences in filters, collimation and softer beam spectrum. Spectral CT offers many benefits by virtue of material-specific information in addition to diagnostic anatomic detail, and VMI low mono-E reconstruction can recover CNR potentially lost with use of higher kVp. To our knowledge, we are the first to explore dose-neutrality in children undergoing abdominal-pelvic CT examinations using single-source duallayer spectral CT. Purpose or Case Report: Determine the spectrum of plain film findings encountered in ileocolic intussusceptions and assess their predictive value in reducibility by therapeutic enema, and in surgical outcomes in patients with failed reductions.

An IRB approved, retrospective study was performed. The radiology information system at our institution was queried for the keyword "intussusception" in fluoroscopic reports from September 2012 -August 2017. Two authors then reviewed plain films obtained prior to therapeutic enema, when available, for the following findings: normal, paucity of bowel gas, soft tissue mass, meniscus sign, or obstruction. When plain films were not available, the scout image from the therapeutic enema was reviewed. Disagreements were resolved by consensus. The medical record was then reviewed for each patient to determine success of therapeutic enema, and surgical outcome in cases of failed reduction. Complicated surgical cases were defined as those that required more extensive surgery than a standard laparoscopic reduction, such as conversion to open laparotomy. Results: 182 total cases of intussusception reduction were identified and reviewed. A normal bowel gas pattern was seen in 14%, paucity of bowel gas in 65%, soft tissue mass in 26%, meniscus sign in 12%, and obstruction in 10%. Percentages total > 100% due to more than one finding in some patients. A normal bowel gas pattern was associated with the highest therapeutic enema success rate (83%) and lowest rates of complicated surgery (4%) and bowel resection (4%). Conversely, bowel obstruction was associated with the lowest enema success rate (21%), and highest rates of complicated surgery (47%) and bowel resection (42%). Patients with bowel obstruction had a significantly higher likelihood of needing surgical reduction (OR 9.0, 95% CI 2.84-28.47) and bowel resection (OR 11.1, 95% CI 3.66-33.86) than those without obstruction. Paucity of bowel gas, presence of soft tissue mass, and meniscus sign showed enema success rates and surgical outcomes intermediate between normal and obstruction. Conclusions: In most cases of ileocolic intussusception, plain films show an abnormality, with paucity of bowel gas being the most common finding. Plain film findings can provide important prognostic information to both the radiologist and the surgeon, with significantly lower enema reduction rates and higher rates of complicated surgeries in patients with bowel obstruction. Results: Age at presentation ranged from prenatal to 16 years with a male to female ratio of 2:1. All patients presented with neck swelling, 13(81%) had neck infection/suppurative thyroiditis at initial presentation. Two patients had severe thyroiditis/mediastinitis requiring ICU admission. Three neonates presented with non-infected cystic neck masses; two had been detected prenatally. Two neonates were presumed to have lymphatic malformations, with spontaneous clinical resolution by three months of age later re-presenting with evidence of PFST. PFST was left-sided in 94%. Eight patients were initially diagnosed with branchial cleft cyst and underwent cyst resection ± hemithyroidectomy without consideration for PFST. These patients re-presented with multiple episodes of recurrent neck abscess. All patients had neck imaging prior to definitive diagnosis. Imaging studies included radiographs, ultrasound, CT, MRI and Upper GI contrast studies. No single modality was diagnostic of PFST in all patients with 75% undergoing multimodality imaging prior to diagnosis. PFST was identified on upper GI study in 6/12 patients. All cases were confirmed by endoscopic visualization. The interval from presentation to endoscopy was 2.6-174 months.Management of PFST was via endoscopic cauterization in 13 patients, surgery in 2, and 1 patient did not require either.

Conclusions: This study highlights the complex nature of PFST. This anomaly is uncommon, has variable clinical and imaging presentations and may have a lengthy, complicated course if not considered at initial presentation. An episode of suppurative thyroiditis in a child should prompt investigation for a PFST. We are the first group to describe the atypical presentation of a neonatal cystic neck mass undergoing spontaneous clinical resolution but re-presenting as a PFST. Purpose or Case Report: Analyze the feasibility of a performing an ultrafast Digital PET/CT on pediatric patients by comparing image quality and SUV quantification of normal structures using multiple simulated time-per-bed position protocols.

This HIPAA compliant retrospective study had IRB approval. 10 patients ranging in age from 3 to 17 years old had an 18 F-FDG PET/CT performed on a digital PET/CT scanner at a tertiary academic hospital. In addition to the clinical standard 120-seconds per bed position scan, simulated PET images were generated using list mode data to create PET data sets at 30-, 60-, and 90-seconds per bed position. Two readers blinded to time-per-bed position assessed the data evaluating image quality using a 5-point Likert scale (1: extremely poor quality study with major artifacts that is not clinically useful, 2: poor quality study with major artifacts, whose clinical use is not advised, 3: average study with moderate artifacts, probably affecting clinical use, 4: good study with only minor artifacts not affecting clinical use, 5: excellent study without artifacts). SUVmax was calculated on seven regions of interest placed on physiologic structures for each data set. Spearman correlation and relative median and interquartile ranges were calculated. Results: 40 reconstructed data sets from 10 patients were analyzed. Mean Likert scores for the 30-second, 60-second, 90second and 120-second images were 3.7, 4.3, 4.6 and 4.7 for Reader 1 and 4.1, 4.9, 4.8, and 4.9 for Reader 2, respectively. Reader 1 found three 30-seconds per bed position scans below clinical acceptance, Reader 2 found all images clinically acceptable. For the 70 regions of interest per data set, there was near perfect SUVmax correlation with the standard 120-seconds per bed position exam (ρ = .98, .994 and .997 respectively for 30, 60 and 90 seconds per bed position, p<.00001). The median and interquartile ranges of the relative SUVmax differences for the 30-, 60-and 90-seconds per bed position exams compared to the clinical standard were 15% (7-27%), 6% (1-14%) and 3% (-1-7%) respectively. Conclusions: Our study demonstrates the feasibility of performing ultrafast PET/CT in pediatric patients on a digital PET/CT. Reduction in scan times of up to 50% (60 seconds) are attainable without adversely affecting the SUVMax measurements or image interpretability. Digital PET/CT using an ultrafast protocol is a promising technology that may result in decreased imaging times and improved patient compliance.

Utility of 18 F-FDG PET/CT following ketogenic diet in detecting endocarditis in children and adult congenital heart disease patients. Purpose or Case Report: Congenital heart disease patients are at increased risk of infective endocarditis, related to implantable devices such as prosthetic valves and pacemarkers. With standard PET/CT protocols, physiologic myocardial glucose uptake limits the assessment of the heart and related cardiac structures. The objective of the present study is to evaluate the utility of 18 F FDG PET/CT following ketogenic diet preparation in children and adult congenital heart disease patients in confirming suspected endocarditis.

Congenital heart disease patients who had undergone FDG PET/CT for suspected endocarditis were identified using a database search. A total of 25 individual patients were identified. Of these, 22 had undergone PET/CT specifically for evaluation of suspected endocarditis. Patients who underwent PET/CT for non-endocarditis indications were excluded from the remainder of the analysis. At the time of their index PET/CT scans, the ages of the 22 evaluated patients ranged from 11 months to 65 years of age, with a median age of 24 years. PET/CT findings were correlated with peripheral blood cultures, surgical findings, explant cultures and histopathology, and clinical course. Results: Ketogenic dietary preparation was successful (myocardial uptake visually suppressed) in 16 of 22 patients. The majority of the patients (18/22) had focal FDG PET/CT findings suspicious for endocarditis, while 4 had negative or nonspecific scans. 14 of 22 patients subsequently were managed surgically; 13 of these patients had explant specimens with histologic evidence of infection and/or positive cultures. The remainder of the patients received antibiotic therapy; five of these clinically managed patients were followed to symptom and/or imaging resolution. Clinically significant non-cardiac PET findings were identified in six patients (e.g. osteomyelitis) that potentially changed management.

Given an adequate index of clinical suspicion, 18 F-FDG PET/CT is a sensitive and specific imaging modality for diagnosing infective endocarditis in the congenital heart disease population, particularly in patients with prosthetic valves or other device implants in whom other imaging modalities such as echocardiography are limited by hardwareassociated acoustic artifacts. In certain patients in whom endocarditis is not diagnosed other sites of remote spread of infection may also be identified. (n=15) intravenous injection of ferumoxytol (5 mg Fe/kg). MRAC was obtained by a two-point Dixon LAVA sequence accounting for fat, air/background, lung, and soft tissue. We compared the signal to noise ratio (SNR), SUVmean and SUVmax, calculated based on body weight (bw) and body surface area (bsa) of blood, brain and normal organs on enhanced and unenhanced MRAC and PET images using a mixed effects linear model. In addition, we compared the number of tissue misclassifications on ferumoxytol-enhanced and unenhanced AC-maps by Poisson regression model. All comparisons were performed assuming alpha of 0.05.

The SNR of the blood, brain, and visceral organs on ferumoxytol-enhanced scans was significantly higher compared to unenhanced scans (p<0.001). The SUVmean and SUVmax values of different organs based on bw or bsa on ferumoxytolenhanced and unenhanced AC-maps were not significantly different (all p>0.05). Ferumoxytol-enhanced AC-maps showed no significant difference in tissue misclassifications (p=0.09). Conclusions: Ferumoxytol administration does not affect SUV measurements on 18 F-FDG PET/MR scans. Therefore, unenhanced scans can be omitted, and the acquisition of PET/MR scans can be accelerated by obtaining all images after primary contrast agent injection. Purpose or Case Report: The tumor microenvironment (TME) in neuroblastoma (NB) is a key regulator of treatment resistance and disease relapse. Myeloid-derived suppressor cells (MDSCs) play a central role in maintaining the TME in NB by suppressing host immunity, driving angiogenesis, and remodeling tissue. Since MDSCs play a central role in angiogenesis, we hypothesized that MDSC-targeted therapies will alter the architecture of tumor vasculature. In this preclinical study, we investigated contrast-enhanced CT (CECT) using a long circulating liposomal-iodine (Lip-I) nanoparticle contrast agent for monitoring changes in tumor vasculature in response to MDSC-targeted immunotherapy.

In vivo studies were performed in three groups of humanized mice: 1) NB tumor cells (Tumor group); 2) NB tumor cells + human MDSCs (Tumor+MDSC group) and 3) NB tumor cells + MDSC cells + Immunotherapy (MDSCdirected Therapy group). For the Therapy group, MDSC-directed natural killer (NK) cells were intravenously administered 10 days prior to CT imaging. CECT was performed at 2 and 4 weeks post-tumor implantation. Lip-I was injected four days prior to CT. On the day of imaging, noncontrast CT was performed to determine tumor vascular permeability. A second dose of Lip-I agent was administered and CT angiogram acquired to determine tumor vascularity. Tumors were segmented to study 3D spatial variations in vascular structure. Animals were then euthanized to determine MDSC burden and peri-vascular localization using flow cytometry and immunohistochemistry (IHC Purpose or Case Report: MRI with gadolinium-based contrast material (GBCM) is one of the modalities used to diagnose and characterize neuroblastoma. However, the use of GBCM is not without risk. Recent data have shown variable retention of GBCM in the brain, bones, and other organs. In addition, GBCM adds cost and time to the imaging study. The purpose of our study was to determine if GBCM are needed to identify image-defined risk factors (IDRF) in patients with neuroblastoma.

A retrospective case-control study was performed. All patients who received an abdominal MRI at the time of diagnosis of intra-abdominal neuroblastoma were included. Each exam was duplicated with the contrast-enhanced images removed from one copy of the exam. A single pediatric radiologist blinded to study purpose reviewed the separated imaging exams in a random order on our research PACS measuring three dimensions of the tumor (width x length x height), reporting the presence or absence of each potential abdominal IDRF and the presence or absence of metastatic disease. Considering the exam including contrast-enhanced images as the reference-standard, Kappa coefficient, sensitivity and specificity were calculated for the detection of IDRF and metastases. A paired t-test was used to compare differences in tumor measurements. Results: 50 patients were included in the study. There was no significant difference between groups for any diameter measurement (p-values: 0.82-0.9). However, there was a statistically significant difference in the detection of each IDRF as well as for the detection of metastatic disease. The kappa value ranged from 0.46 to 0.94 for IDRF detection (p-value range 0.007 to <0.001) and from 0.23 to 1 for the detection of metastatic disease (p-value range 0.01 to <0.001). The sensitivity of IDRF detection ranged from 0.81 to 1 while specificity ranged from 0.33-1. The highest sensitivity of the non-contrast images was for detection of lower mediastinal invasion, superior mesenteric artery encasement, and lung metastasis while the highest specificity was for infiltration of the porta hepatis, encasement of the iliac vessels, invasion to the liver, and detection of lung metastases. Conclusions: GBCM does not significantly improve a radiologist's ability to measure neuroblastoma size. However, use of GBCM is associated with a significant difference in the detection of IDRFs at diagnosis.

The Purpose or Case Report: Determine radiologists' and oncologists' perceived impact of the structured report on clinical workflow, point of care teaching, and research in a large academic pediatric institution.

Fellows and attending radiologists from both radiology and oncology departments at our institution completed an online survey regarding their perceptions of the impact of the structured report on workflow, teaching and research. Surveys were sent to 42 radiologists and 51 oncologists.Two different surveys were created with specific directed questions for radiology and oncology; there were 7 questions in the oncology survey and 10 in the radiology survey. The structured report template was included with the survey to serve as a convenient reference. Results: The survey was completed by 50% of the radiologists and 27% of the oncologists. All oncologists claimed their preference of structured over free text reports. 70% of oncologists reported fewer questions for the radiologists after implementation of structured reporting. 64.3 % reported improved workflow efficiency and improved clinical management of their patients. 75% of radiologists preferred structured reporting, with 70% claiming improved accuracy of their reports and a more efficient process when asked to review cases with the referring clinicians. Of interest, 80% of radiologists considered using structured reports as a teaching tool while 50% of the oncologists used them for this purpose Conclusions: Our results show that radiologists and oncologists perceive a significantly positive impact of the structured report for pediatric tumors on workflow and efficiency of initiating care. In addition the structured report provides a teaching tool to enhance the current educational experience for trainees in both radiology and oncology programs. Purpose or Case Report: To investigate if Ferumoxytol (Fe)enhancement in MRI scans influences assessment of tumor size and extension in pediatric bone tumors compared to Gadolinium (Gd)-enhancement in a clinically relevant manner. We hypothesized that tumor size and extension can be equally assessed on Fe-MRI scans and Gd-MRI scans.

In this retrospective, IRB-approved study, we compared quantitative and qualitative image parameter of bone tumors in 13 patients (11 osteosarcomas, 2 Ewing sarcomas; 9m/4f; 15.9 ± 4.7y/o) that received a Fe-and a Gd-MRI scan within one month of each other and before treatment initiation. One investigator measured the maximum tumor length in 3 orientations and the tumor volume on T1w-LAVA, T1w-SE and T2w-FSE sequences. Results were compared using Bland Altman analysis. In addition, 3 radiologists independently evaluated tumor involvement of the diaphysis, metaphysis and epiphysis using a Likert-scale. Results were pooled across readers and then analyzed according their agreement on tumor assessment (affectedyes/no/uncertain) and confidence (uncertain/probably/definitely). Results: Tumor lengths were not significantly different on Fe-MRI and Gd-MRI scans with mean differences of 0.13±0.72cm (axial-short axis), 0.28±0.59cm (axial-long axis) and 0.29±0.79cm (coronal) with 95% limits of agreement (95%LoA) of -1.27-1.53cm, -0.88-1.43cm and -1.84-1.26cm respectively and high concordance correlation coefficients (CCC) of 0.97 (axial-la) and 0.98 (axial-sa/cor). The tumor volume also showed a strong similarity with a mean difference of 5.86±43.81 mm 2 with 95%LoA of -80.0 -91.7 mm 2 and an even higher CCC of 0.99. The assessment of tumor presence in the metaphysis and the associated confidence showed perfect agreement for all sequences (1.0, 1.0) of the Fe-and Gd-scans. For the diaphysis, the assessment and confidence, respectively, showed highest agreement for T1-LAVA (0.99, 0.98) and lower agreement for both T2-FSE (0.87, 0.85) and T1-SE (0.87, 0.85). For the epiphysis, the assessment and confidence showed the highest agreement for T2-FSE (0.94, 0.90), slightly lower for T1-LAVA (0.91, 0.85) and the lowest for T1-SE (0.77, 0.73).

Our results suggest that Fe-enhanced MRI scans can be equally used for assessment of tumor size and extension in pediatric bone tumors compared to Gd-enhanced MRI scans. Purpose or Case Report: To assess the accuracy of imaging techniques currently used in practice for early detection of tumors in patients with Li-Fraumeni syndrome (LFS).

Two radiologists retrospectively reviewed all available imaging data from patients with LFS (January 1999 to September 2017). Based on the results from the imaging analysis, lesions were categorized into benign, malignant and indetermined. Histopathological and clinical information were also obtained and independently reviewed. Reference standard measures were (1) findings in dedicated cross-sectional imaging obtained within 6 months from index scan;

(2) histopathologic diagnosis; or (3) clinical outcomes obtained at least 3 years from the index scan. Imaging accuracy was defined by the true positive (TP), true negative (TN), false positive (FP) and false negative ( Purpose or Case Report: To evaluate region-specific perfusion alterations in neonatal hypoxic ischemic injury (HII) based on pulsed arterial spin labeling (ASL) magnetic resonance imaging (MRI) data as compared to controls.

The ASL perfusion data of 19 normal neonates (10 male/9 female, 12 ± 6 days) and 31 HII neonates with positive MRI findings (14 male/17 female, 9 ± 7 days) were retrospectively evaluated. The cerebral blood flow (CBF) maps were first calculated using the ASL data processing toolbox (ASLtbx). From the CBF calculation results, averaged region-specific perfusion values were extracted for comparison. Specifically, the JHU neonate atlas was used to specify 130 brain regions for perfusion quantification. For accurate comparison of each brain region, both CBF image and JHU atlas were aligned to the subject-specific T2-contrast structural image space. Permutation test was used to identify brain regions with significant perfusion alterations between the control and HII group with positive MRI findings.

The experimental results identified 10 brain regions significantly different in perfusion between the control and HII groups based on the permutation test with p<=0.01.The brain regions with the most significant perfusion alterations included left/right precuneus (p=0.005/0.007), left/right superior occipital gyrus (p=0.002/0.004), left/right cuneus (p=0.001/0.001), left/right lingual gyrus (p=0.005/0.005), left/right cerebellar hemisphere (p=0.001/0.003). Other than the region-specific perfusion comparison, the whole brain comparison was also implemented between the two groups by stacking all 130 average values of each brain region according to the JHU atlas.

The results also showed the significant difference in whole brain perfusion (p<0.0001) with the average and standard deviation value (ml/100g/min) to be 11.5±4.7 and 17.1±7.3 for the control and HII groups, respectively.

The region-specific perfusion analysis based on the brain ASL data can help identify brain regions most significantly affected by the hypoxic ischemic injury. Further studies are warranted to explore the prognostic implications of perfusion alterations in these regions.

Changes in brain perfusion in successive arterial spin labelling MRI scans in neonates with hypoxic-ischemic encephalopathy Purpose or Case Report: The primary objective of this study was to evaluate changes in cerebral blood flow (CBF) using arterial spin labeling MRI between day 3 of life (DOL3) and day 10 of life (DOL10) in neonates with hypoxic-ischemic encephalopathy (HIE) treated with hypothermia. The secondary objectives were to compare CBF values between the different regions of interest (ROIs) and between infants with ischemic lesions on MRI and infants with normal MRI findings.

We prospectively included all consecutive neonates with HIE admitted to the neonatal intensive care unit of our institution who were eligible for therapeutic hypothermia. Each neonate systematically underwent two MRI examinations as close as possible to day 3 (early MRI) and day 10 (late MRI) of life. A custom processing pipeline of morphological and perfusion imaging data adapted to neonates was developed to perform automated ROI analysis. Results: Twenty-eight neonates were included in the study between April 2015 and December 2017. There were 16 boys and 12 girls. Statistical analysis was finally performed on 37 MRIs, 17 early MRIs and 20 late MRIs. Eleven neonates had both early and late MRIs of good quality available. Eight out of 17 neonates (47%) had an abnormal early MRI and 7/20 neonates (35%) had an abnormal late MRI. CBF values in the basal ganglia and thalami (BGT) and temporal lobes were significantly higher on DOL3 than on DOL10 (p<0.05). There were no significant differences between DOL3 and DOL10 for the other ROIs. CBF values were significantly higher in the lobe grey matter (GM) vs. the brain and in the BGT vs. the cortical GM, on both DOL3 and DOL10 (p<0.05). On DOL3, the mean CBF was significantly higher in the cortical GM, the BGT, and the frontal and parietal lobes in subjects with an abnormal MRI compared to those with a normal MRI (p<0.05). The perfusion differences between subjects with an injury and those without an injury on MRI had disappeared on the DOL10 scan.

Conclusions: This study is, to our knowledge, the first to evaluate CBF on two successive scans within the first 15 days of life in the same subjects in this clinical setting. ASL imaging in asphyxiated neonates therefore seems more relevant when used relatively early, in the first days of life. The correlation with neurodevelopmental outcome warrants investigation in a large cohort, to determine whether CBF values can provide prognostic information beyond that provided by visible structural abnormalities on conventional MRI. Purpose or Case Report: To develop a quantitative method for detection of neonatal hypoxic ischemic injury (HII) based on pulsed arterial spin labeling (PASL) magnetic resonance imaging (MRI) perfusion data which can be served as a reference standard for brain contrast enhanced ultrasound (CEUS), a perfusion based ultrasound technique.

The ASL perfusion data of 19 controls (10 male/9 female, 12 ± 6 days) and 31 neonates with HII and positive findings on MRI imaging (14 male/17 female, 9 ± 7 days) were identified for retrospective analysis. After calculating the cerebral blood flow (CBF) map from the ASL data using the ASL data processing toolbox (ASLtbx), the quantitative ASL perfusion ratios of central gray nuclei to white matter (GNW) and central gray nuclei to cortex (GNC) were compared between the control and HII groups by permutation test. The perfusion ratios were calculated on the mid-coronal plane as defined by the plane in which maximum cross sectional area of the central gray nuclei is visualized. The mid-coronal slice has previously been used for the wash-in time intensity curve on brain CEUS, and the selection of this slice for ASL quantification was based on the premise of developing a reference standard for developing brain CEUS for diagnosis of neonatal HII.

Our results demonstrate that quantitative perfusion ratios of GNW and GNC are significantly different between the control and HII groups. Based on the permutation test, the white matter of cingulum hippocampal part (p=0.013) and cerebral peduncle (p=0.004) showed significant difference by GNW comparison, and the cortex of superior temporal gyrus (p=0.045), middle temporal gyrus (p=0.010), inferior temporal gyrus (p=0.043), parahippocampal gyrus (p=0.006), hippocampus (p=0.016), and insular cortex (p=0.008) showed significant difference by GNC comparison. The ratios of central gray nuclei to entire white matter, entire cortices, and all brain regions on the middle coronal plane excluding gray nuclei were also extracted for comparison, and the results revealed significant group differences with p values of 0.035, 0.017, and 0.013, respectively. Conclusions: Quantitative perfusion ratios can be used to detect the presence of brain injury, which has critical implications for developing perfusion-based ultrasound techniques including brain CEUS. The utilization of quantitative perfusion ratios also minimizes inter-subject and technical variability while allowing subject's own perfusion as internal control. Purpose or Case Report: Contrast-enhanced ultrasound (CEUS) is a new and promising modality for evaluating neonatal brain injury through the detection of alterations in perfusion. Qualitative and quantitative evaluation of perfusion can be performed for diagnostic interpretation of brain CEUS scans. As brain CEUS is a novel application, there are challenges in implementing this into the clinical setting. We explored whether a brief didactic session on the normal brain CEUS perfusion pattern can improve radiologists' interpretation of brain CEUS scans.

Four attending pediatric radiologists evaluated 6 CEUS mid coronal view cine clips of neonatal brains demonstrating either normal perfusion or various patterns of perfusion abnormalities over a two-day period. Findings on all brain CEUS scans were confirmed by magnetic resonance imaging. On a provided questionnaire, evaluators indicated if each example was normal or abnormal (diagnosis). If an abnormality was deemed present, evaluators specified the distribution (focal, multifocal, diffuse), location (right hemisphere, left hemisphere, bilateral hemispheres) and perfusion characteristics (hypoperfused, hyperperfused). After initial evaluation, a brief didactic session was provided, demonstrating the normal neonatal brain perfusion pattern. On the following day, participants reevaluated the same cases, using the same questionnaire. Distribution of error pre and post education was analyzed using McNemar testing for responses in each of the following categories: diagnosis, distribution, location and perfusion.

The perfusion category showed 10 incorrect responses pre education and 4 incorrect responses post education, p=0.07. The diagnosis category showed 8 incorrect responses pre education and 2 incorrect responses post education, p=0.07. The location category showed 11 incorrect responses pre education and 6 incorrect responses post education, p=0.18. The distribution category showed 13 incorrect responses pre education and 9 incorrect responses post education, p=0.22. Conclusions: Preexisting familiarity with intravenous contrast use and education on the normal neonatal brain perfusion pattern on CEUS was insufficient to significantly alter exam interpretation, although a trend toward improvement in all areas was observed. This suggests that more familiarity with the concept is warranted and that there is a potential benefit in developing quantitative methods to be used in addition to qualitative methods for detecting injury on brain CEUS. Purpose or Case Report: Advancements in contrast-enhanced ultrasound (CEUS) have allowed analysis of perfusion metrics in various organs, including the brain. The influence of hemodynamic parameters such as heart rate (HR), mean arterial pressure (MAP), and blood flow have not been delineated. The goal of this study was to assess for correlation between hemodynamic parameters and perfusion metrics in CEUS of the brain in fetal sheep maintained on the EXTra-uterine Environment for Neonatal Development (EXTEND) system.

Following IACUC-approved protocols, 9 premature fetal lambs were transferred from placental support to the EXTEND system. 0.1-0.3 mL Lumason® ultrasound contrast was administered into the pumpless umbilical arterial to umbilical venous oxygenator circuit at varying gestational ages (93-133 days). Images were acquired as 90 second cine clips using a GE Logiq E9 ultrasound system and C2-9 transducer with settings optimized for contrast visualization. Clips were analyzed with VueBox TM post-processing software and timeintensity-curves (TICs) were generated. Hemodynamic parameters including HR, MAP, blood flow through the oxygenator, pre-oxygenator oxygen level, oxygen delivery, preand post-oxygenator pressure differential, and sweep requirements across the oxygenator were recorded continuously and averaged over the 90 seconds of imaging.

Results: A total of 86 CEUS examinations were performed, 72 of which were quantifiable and included in analysis. A multilevel mixed-effects linear regression was performed with random intercepts by subject. Hemodynamic parameters had no effect on mean transit time (MTT), rise time (RT), fall time (FT), or time-to-peak (TTP) (p>0.06). Sweep requirement and oxygen delivery demonstrated an association with wash-in rate (WIR), wash-out rate (WOR), wash-in area-under-the-curve (WiAUC), wash-out AUC (WoAUC), wash-in-wash-out AUC (WiWoAUC), and wash-in-perfusion-index (WiPi) (p<0.03). Conclusions: CEUS perfusion metrics dependent on time including MTT, RT, FT, and TTP are not correlated with systemic hemodynamic data and any detectable changes in these perfusion parameters reflects differences in localized blood flow. None of the parameters need to be corrected for HR or MAP. Perfusion parameters dependent on AUC quantification including WiAUC, WoAUC, WiWoAUC, and WiPi are correlated with sweep requirement and oxygen delivery and these associations should be taken into consideration when interpreting TIC results in future studies. Purpose or Case Report: To evaluate the feasibility of artificial intelligence detection of germinal matrix hemorrhage (GMH) on neurosonograms using convolutional neural networks (CNN).

In this HIPAA-compliant, institutional review board-approved study, a single-institution radiology report database was queried to identify neonatal head ultrasounds (US) performed between January 2016-July 2018, which were then categorized according to presence/absence as well as the grade (I-IV based on the Papile classification) and side of GMH. Static US images (gray-scale coronal and sagittal views from the anterior fontanelle) from these exams were used to train, validate, and test the CNN. From a total of 1000 images, 800 (n=400 each) served as the training dataset, 100 (n=49 for normal and n=51 for abnormal) as the validation dataset, and 100 (n=51 for normal and n=49 for abnormal) as the test dataset. Performance of the CNN was determined by calculating sensitivity, specificity and accuracy based on the clinical radiology report reference. Area under the receiveroperating curve (AUC) was also obtained. Results: A total of 387 US exams were identified, including 325 normal exams and 62 that were positive for GMH. The mean corrected gestational age at the time of imaging for normal studies was 38 weeks 3 days, compared with 30 weeks 6 days for studies positive for GMH. In the 100 case test set, the overall accuracy of CNN was 72% for GMH detection with an AUC of 0.79. 44/51 normal cases were correctly classified (specificity=86.3%). However, sensitivity was low (57.1%) with a high false negative rate (28/49). Conclusions: Artificial intelligence is currently not able to detect GMH on head ultrasound examinations with acceptable accuracy, primarily due to low sensitivity. This suggests that ultrasound images may be less amenable to CNN training than CT or radiographs due to decreased image contrast, and reaffirms the primary role of human radiologists in head ultrasound interpretation. Recently, high resolution chest ultrasound of the lung surface has been proposed as a bedside method for diagnosing pulmonary lymphangiectasia (PL). We compare high resolution chest ultrasound of the pleural surface to DCMRL in patients with suspected PL.

We retrospectively reviewed the high resolution chest ultrasound in children with suspected PL who also had DCMRL. Transverse and coronal images in the upper, middle and lower lung surfaces along the mid-axillary and midclavicular lines were acquired on each side using a high frequency (15mHz) linear transducer. The presence or absence of lung surface irregularity, sub-pleural cysts, and pleural effusions was documented. Correlation was made with intranodal lymphatic contrast perfusion to the lung interstitium or pleural space on DCMRL and the presence of interstitial edema and pleural effusion on heavily T2 weighted imaging. Results: 5 children (4 boys) between the ages of 1 month and 4 years old underwent both high resolution chest ultrasound and DCMRL. Chest ultrasound and DCMRL were performed in average 4.4 days apart (range 1 to 7 days). Four patients had congenital heart disease and one had a primary lymphatic disorder.4 of 5 patients had sonographic findings of PL including lung surface irregularity (3 bilateral, 1 unilateral, 1 absent) and sub-pleural cystic spaces (1 bilateral, 2 unilateral, 2 absent). All 5 patients had pleural effusions (3 bilateral, 2 unilateral) on ultrasound. Similarly, at DCMRL 4 of 5 had abnormal pulmonary lymphatic perfusion (3 bilateral, 1 unilateral, 1 absent). Interstitial edema (5 bilateral) and pleural effusions (4 bilateral, 1 unilateral) were present on T2 weighted imaging. US and DCMRL findings were in agreement in all 5 patients. Lung surface irregularity and sub-pleural cysts on ultrasound correlated with abnormal pulmonary lymphatic perfusion on DCMRL Conclusions: High resolution chest ultrasound of the pleural surface shows sonographic findings that correlate with findings on DCMRL in patients with pulmonary lymphangiectasia. US seems to be a promising modality for the screening of PL in children to avoid invasive DCMRL. Purpose or Case Report: To evaluate the diagnostic accuracy of contrast enhanced ultrasound (CEUS) for the evaluation of thoracic duct (TD) patency following percutaneous intranodal ultrasound contrast injection into the groin.

CEUS examinations of the left neck following intranodal injection of ultrasound contrast (sulfur hexafluoride lipid-type A microspheres) were retrospectively evaluated for TD outlet patency by 2 blinded and independent pediatric radiologists with experience in both lymphatic and CEUS imaging. Inguinal lymph node access was obtained under ultrasound guidance. CEUS exams were performed concurrently with intranodal inguinal injection of contrast and imaged continuously for 3-5 minutes or until contrast appeared. The TD outlet was classified as patent or not patent by visualization of lymphatic contrast passage into the venous system. Results were correlated with conventional lymphangiography (when available) which was reviewed independently and blinded by a lymphatic interventionalist. Results: 11 patients (9 male) who presented for lymphatic evaluation were identified for the present study. Mean age was 4.4 years (range 2 months to 13.8 years). Of the 11 cases, the TD outlet was patent on CEUS in 10 patients and not patent in one. There was agreement between both CEUS readers on all cases. 9 of 11 patients also had conventional lymphangiography (8 of 9 on the same day, 1 within 6 months of CEUS). Of these 9 cases, the TD outlet was patent in 8 patients and not patent in one on conventional lymphangiography. These patency results matched the CEUS findings for each patient, thereby demonstrating agreement in all 9 cases. Conclusions: CEUS of the neck following intranodal inguinal ultrasound contrast injection can accurately determine TD outlet patency and can thus eliminate the need for conventional lymphangiography when no percutaneous intervention is needed. Though dose was higher for DECT in some larger patients, this may be due to small sample size. Regardless, DECT has several advantages over SECT, which may outweigh a potential small increase in radiation dose on a case by case basis. We suggest that DECT can be introduced into routine practice for pediatric thoracic CTA. 

We evaluated multiple phases of the free breathing (sedated/non-sedated) dynamic 4D airway CT angiograms performed for investigation of tracheobronchomalacia on a dual energy CT scanner with extended craniocaudal dynamic phase coverage below L1. Image acquisition was initiated with peak descending aortic enhancement (425 HU). Qblique coronal 1.3-1.5 mm thick maximum intensity projections were reconstructed from the phase where the AoA was identified as a hairpin curved vessel in the anterior midsagittal surface of the spine. The AoA was considered visualized if the "hairpin loop" of the artery within the spinal canal was visible in addition to the anterior spinal artery and the vessel could be traced to the intercostal artery by paging through oblique coronal reformats. The vertebral body above the level at which the spinal branch of the segmental artery entered the spinal canal and continued as the AoA was considered the level of AoA origin. The study radiologist blinded to the official report reviewed all scans. Results: Dynamic 4D airway CTs in 79 patients (43 male, 5 months-18 years) who underwent preoperative imaging for airway anomalies were evaluated. Continuity between AoA and aorta through intercostal or lumbar artery was confirmed in 67 (84%) patients. Level of origin of AoA was most frequent at T10 and T11 (19 and 24 patients, respectively) but was also found as high as C7. AoA was most commonly seen on 3 rd acquisition (56%). Of 12 patients in whom all the criteria for AoA were not fulfilled, the anterior spinal artery was visualized in the lower spinal canal in 5 patients. Conclusions: AoA can be identified in pediatric patients on dynamic 4D airway studies with greater craniocaudal coverage, optimal timing of scan acquisition and tailored multiphase data processing. This information is valuable to surgeons and helps avoid more invasive procedures like DSA. Purpose or Case Report: Hereditary hemorrhagic telangiectasia (HHT) is a genetic disorder. Pulmonary AVM (PAVM) is one of the classic manifestations and may be progressive in nature.The purpose of this study is to characterize CT angiographic (CTA) findings of PAVM in children with HHT and to correlate these with graded transthoracic contrast echocardiography (TTCE).

A total of 40 patients (median age 14.9 years) with a diagnosis of HHT who had undergone both CTA and TTCE were included. With CTA, PAVM was evaluated based on location, distribution, and size. Each PAVM was scored with a grading system being applied to conventional and maximum intensity projection (MIP) images as follows:0 = nodule, but unlikely AVM, 1 = ground glass opacity (GGO), 2 = GGO with increased vascular network, 3 = GGO or nodule with single draining vein, 4 = GGO or nodule with equally sized afferent and efferent vessels, 5 = GGO or nodule with afferent and asymmetrically enlarged efferent vessels, 6 = true AVM with aneurysmal sac without the nodule or GGO. Total number of PAVM, cumulative PAVM grading (adding up individual PAVM grades), the highest grade given, total volume, and age were recorded for each patient. TTCE was graded as per Gazzaniga et al based on the number of bubbles (0 to 3) from pulmonary right to left shunt and then correlated with the variables of CTA and patient's age (Pearson correlation). Results: A total of 124 pulmonary lesions were identified on CTA, including; peripheral (n=112), central (n=10), and both (n=2); all the lesions were localized without diffuse form. The CTA grading was increased on MIP images in 39 out of 124 PAVMs on conventional CT. The distribution of grades on MIP included grade 0 (n=9), grade 1 (n=16), 2 (n=9), 3 (n=42), 4 (n=34), 5 (n=2), and 6 (n=12). The median size and volume were 3.9mm and 31.1mm 3 respectively. A statistically significant correlation was seen between all variables of CTA and TTCE (p<0.05), with the strongest correlation seen in those with the CTA highest grading (r=0.79, p<0.0001). Age did not correlate with any variables of CTA or TTCE grading. Conclusions: In children with HHT, GGO or nodule with single (grade 3) or two vessels (grade 4) were seen in two thirds of cases. MIP images enabled more precise detection of the afferent or efferent vessels. The highest grade in each patient had the strongest correlation with the severity of pulmonary right to left shunt. Unlike in adults, age was not an important factor in determining the severity of PAVM in children. Purpose or Case Report: Pediatric MRI often involves sedation or general anesthesia (S/GA) to minimize bulk and respiratory motion to improve image quality, especially for uncooperative patients. To minimize the duration and intensity of S/GA and to potentially eliminate S/GA completely, rapid imaging is essential. In this work, we enable rapid pediatric chest imaging with a time-efficient 3D cones trajectory and deep-learning off-resonance artifact correction.

A prospective study of 30 pediatric chest magnetic resonance angiography exams (16 female; mean age, 5.9 years; range, 0-17 years) was performed to train and validate a residual convolutional neural network to correct offresonance artifacts (Off-ResNet). Each exam acquired a shortreadout scan (1.18msec ±0.38) and a long-readout scan (3.35msec ±0.74) at 3T. Short-readout scans, with longer scan times but negligible off-resonance blurring, were used as reference images and augmented with additional off-resonance for supervised training examples. Long-readout scans, with greater off-resonance artifacts but shorter scan time, were corrected by autofocus and Off-ResNet and compared to shortreadout scans by normalized root-mean-square error (NRMSE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR). Scans were also compared by scoring on eight anatomical features by two radiologists, using analysis of variance with post-hoc Tukey's test. Reader agreement was determined with intraclass correlation. Results: Long-readout scans were on average 59.3% shorter than short-readout scans (P<0.001). The proposed method had superior NRMSE, SSIM, and PSNR compared to uncorrected images across ±1kHz off-resonance (P<0.01). The proposed method had superior NRMSE over -677Hz to +1kHz and superior SSIM and PSNR over ±1kHz compared to autofocus (P<0.01). Radiologic scoring demonstrated that long-readout scans corrected with Off-ResNet were non-inferior to shortreadout scans (P<0.01).

The proposed method enabled scan time reduction by 59.3% with longer readout durations, while maintaining noninferior image quality to diagnostically-standard scans, suggesting viability of accelerating 3D cones trajectories for pediatric imaging. Additional acceleration can be achieved by combining parallel imaging and compressed sensing, both of which pair well with acceleration from longer readout durations. . Severe complications were encountered in 13 patients (57%) and included fluid collection (n=11), tracheoesophageal fistula (TEF) (n=3), vocal cord paralysis(n=1), and discitis(n=1). There were no patients with vascular injury (0). TEF on MRI was graded as negative/possible/probable with a sensitivity of 100% and specificity of 85% vs. bronchoscopy. Location of the battery was cervical esophagus in all but 1 patient with severe complications (92%). Time from ingestion of BB did not correlate with complications (p=0.65). All complications were visualized on initial MRI performed in first 3 days following retrieval. Subsequent serial examinations did not demonstrate new latent complications (>1 week post retrieval) with a trend to decreased mediastinal inflammation on serial imaging.

Conclusions: MRI after BBI demonstrates a high diagnostic ability to identify the common associated complications including fistula, fluid collection, and TEF. No conclusion can be drawn regarding vascular injury risk as no patients in this cohort had vascular injury. However, in clinical practice, the detection of inflammation receding from the vascular structures on serial MRIs provides invaluable information utilized to make important management decisions. To our knowledge, this is the first paper to report specifically on the MR findings after BBI. Purpose or Case Report: Assess survival and pulmonary outcomes in a cohort of fetuses with severe oligohydramnios who underwent maximal neonatal resuscitative efforts.

We evaluated 33 fetuses with severe oligohydramnios, from any cause, seen in the prior 5 years at either the Boston Children's Hospital Advanced Fetal Care Center or Brigham and Women's Hospital Center for Fetal Medicine, whose parents were counseled on risk of lethal pulmonary hypoplasia based on prenatal sonographic findings and who chose to have maximal neonatal resuscitative efforts at birth. These efforts included immediate intubation, high frequency ventilation, chest tube evacuation of pneumothoraces and inhaled nitric oxide for pulmonary hypertension. A retrospective medical record review compared fetal imaging biomarkers associated with pulmonary hypoplasia (including gestational age (GA) at onset of oligohydramnios, AFI, presence/appearance of kidneys, presence of urine in the bladder, appearance of lungs on MRI) in surviving and nonsurviving newborns. Results: Standard imaging biomarkers did not reliably predict poor outcome. Oligohydraminos was defined as AFI ≤ 5 cm or deepest vertical pocket (DVP) < 2 cm; anhydramnios was defined as absent amniotic fluid. Median GA at birth was 35.5 weeks. Intubation in the delivery room was required in 50% and pneumothorax occurred in 50% of subjects. 77% of the cohort survived to discharge home. Only 3/8 fetuses with anhydramnios survived. Mean GA at onset and duration of oligohydramnios did not differ significantly between survivors and non-survivors. 6/7 deaths occured in the delivery room; 1 NICU death occured at 12 days. 6/7 subjects who died had autosomal recessive polycystic kidney disease (ARPKD), all of whom had large echogenic kidneys; however, 9/16 other fetuses in this cohort also had ARPKD with large echogenic kidneys and survived. 6/7 subjects who eventually had a renal transplant did not require peritoneal dialysis. One subject is on dialysis awaiting transplant. Of 2 infants with a prenatal diagnosis of renal agenesis, one was found postnatally to have tiny dysplastic kidneys, and survived to transplant. Of 7 infants with no fluid in the bladder in utero, only 1 survived. Conclusions: Standard imaging biomarkers did not reliably predict survival. Fetuses can survive what is predicted to be "lethal" oligohydraminios if aggressive neonatal resuscitation takes place and renal replacement therapy is initiated. More hopeful counseling than is currently presented may be warranted. Purpose or Case Report: Given the dramatic structural changes of fetal neurodevelopment, precise knowledge of gestational age and the associated neuroanatomical hallmarks is key to the accurate interpretation of the fetal brain MRI. In this study, we aim to develop and validate a novel deep learning approach to accurately predict gestational age from fetal brain MRI.

We retrospective reviewed all fetal MRIs acquired by our institution from 2004-2017. Normal brain development was assessed by a rigorous review of each study firstly by review of the initial radiology report followed by further evaluation of the fetal brain images by an expert pediatric neuroradiologist. Gestational age was calculated using the current obstetric standard of care approach that references the estimated date of delivery recorded at the first-trimester ultrasound. The robust database of 855 normal fetal brain MRIs was distributed across training (70%), validation (10%) and testing (20%) sets for experimental evaluation. We developed a novel attention-aware deep learning network consisting of two convolutional neural networks (CNNs) working in parallel to combine whole slice image features with brain-specific features, derived from engineered attention maps. This innovative approach mimics the expert neuroradiologists' behavior, allowing a rotation-invariant auto-detection of the fetal brain and thus boosting the prediction outcomes. Regression model performance was evaluated by the R2 statistic. Results: Our attention-aware model achieved strong predictive performance (R2 = 0.93, Mean error = 7.4 days) of gestational age using multi-plane fetal MRI. Single-plane prediction of gestational age achieves an R2 of 0.88 (axial), 0.87 (coronal), and 0.83 (sagittal) with attention-map guidance. A standard CNN without attention enhancement was able to retain high R2 of 0.83, 0.84, and 0.81 for the corresponding axial, coronal, and sagittal MRI planes, respectively. Conclusions: Deep neural networks can accurately characterize healthy in utero brain development from fetal MRI. Clinical translation of such insights has the potential to improve early detection of normal and abnormal fetal brain maturation. The proposed deep learning algorithm could serve as an adjunct diagnostic tool in evaluating appropriate fetal neurodevelopment on MRI. including microtia (n = 9); low set ears (n = 17); external auditory canal malformation or atresia (n = 9); mass involving or abutting the external ear (n = 4, most commonly lymphatic malformations); preauricular skin tags (n = 3); and anotia (n = 2). The most commonly observed inner ear malformations included cochlear malformation (n = 6); absence and / or malformation of the semicircular canals (n= 4); vestibular malformation (n = 3); malformed internal auditory canals (n = 3); and absence of the inner ear structures (n = 1). Ear abnormalities were associated with syndromes on the basis of additional imaging findings, genetic testing and / or postnatal examination in a large number of cases (18/41) with the most commonly observed syndromes including CHARGE (n = 4), Trisomy 18 (n = 3), Trisomy 13 (n = 3), 22q11.2 duplication (n = 2), Brachio-Oto-Renal syndrome (n = 1), Trisomy 21 (n = 1), Trisomy 22 (n = 1), Goldenhar syndrome (n = 1), Klinefelter Syndrome (n = 1), and Cat-Eye Syndrome (n = 1). Syndromic associations were suggested in 5/41 cases, but were not confirmed due to redirected care. Vascular anomalies counted for 4/41 cases involving the external ear. In the remainder of the ear malformations, no syndromic correlation, genetic abnormality or unifying diagnosis was made. Conclusions: Our findings suggest that detailed evaluation of the inner and external ears should be performed in every evaluation of the fetus and malformation may suggest the presence of underlying syndromic condition. Purpose or Case Report: Greater than 60% of placentae from low birth weight infants show signs of hypoxic or ischemic injury from vascular hypo-perfusion. Placental fractional blood volume (FBV) is indicative of perfusion and may be used as a marker of local ischemia. Non-invasive methods for the estimation of placental FBV are therefore of interest in the study of placental pathologies. In this pre-clinical study, we investigated contrast-enhanced magnetic resonance imaging (MRI) for the estimation of placental FBV in a pregnant mouse model. A high T1 relaxivity blood-pool liposomal-gadolinium (liposomal-Gd) contrast agent, which does not permeate placental barrier in rodents, was used for the determination of placental FBV.

In vivo studies were performed in pregnant C57BL/6 mice (8-10 feto-placental units per dam). MRI was performed on a 1T scanner at day 18 of gestation. Precontrast and post-contrast images were acquired using a T1weighted 3D gradient-recalled echo sequence. Post-contrast images were acquired following intravenous administration of liposomal-Gd (0.1 mmol Gd/kg). A variable flip-angle method was used to determine T1 relaxation time and relaxation rates in the placenta (R1 P =1/T1 P ) and the inferior vena cava (R1 IVC =1/T1 IVC ). Differences between pre-contrast and postcontrast R1 values were determined in the placenta (ΔR1 P ) and IVC (ΔR1 IVC ). MRI-derived placental FBV was calculated as a ratio of R1 differences in the placenta to IVC:

where signal is proportional to concentration of iodine contrast agent, was used for validation of MRI-derived FBV. CT scans were performed on a small animal micro-CT scanner. CECT was performed after intravenous administration of a liposomal-iodinated agent (1.1 g I/kg). CT-derived placental FBV (FBV CT ) was calculated as the ratio of signal enhancement in placenta to IVC.

The long circulating property of liposomal-Gd resulted in uniform vascular signal enhancement. T1 relaxation time reduced nearly four-fold in the IVC (pre-contrast T1 IVC = 1230 ± 50 ms; post-contrast T1 IVC = 310± 25 ms). In the placenta, T1 relaxation times decreased three-fold (pre-contrast T1 P = 1870 ± 150 ms; post-contrast T1 P = 540 ± 70 ms). MRI-derived placental FBV was computed as 0.55 ± 0.07 and showed good agreement with values derived from CT (FBV CT = 0.52 ± 0.03).

Conclusions: Contrast-enhanced MRI using a liposomal-Gd blood-pool contrast agent enables accurate determination of placental fractional blood volume. Purpose or Case Report: The purpose of this study was to gain an understanding of the distribution, quantity, and associations of growth recovery lines (GRLs) in children from 0 to 24 months of age with high and low risk for child maltreatment.

We performed a retrospective cohort study of children from 0 to 24 months who had skeletal surveys and an assessment regarding the level of concern for maltreatment. Two pediatric radiologists blinded to the abuse likelihood independently counted the number of GRLs at each proximal and distal bone site of the extremities. A GRL was defined as a radiodense band traversing parallel to the metaphysis involving at least 50% of the width of the metaphysis. Subjects were classified into 1 of 3 groups: low risk (no findings of abuse or neglect on examination), physical abuse, and neglect. Assessments were conducted by at least one child abuse pediatrician prior to the radiologists' review. The demographic data (age in months and gender) and the key outcomes (total number of GRLs in each subject and number of bone sites with GRLs) were tested for association with the risk group variable using Chi-squared and ANOVA. Results: Of the 135 children in this study, 58 were in the low risk group, 26 were in the neglect group, and 51 were in the physical abuse group. Children in the neglected and physically abused groups had 1.73 (p = 0.007) times and 1.84 (p < 0.001) times more GRLs than the low risk group, respectively. The specificity for maltreatment in subjects with at least 10 GRLs in the long bones was greater than 84% [range 84% to 95%], while the sensitivity was less than 35% [range 25% to 35%]. The most common location for GRLs in abused children was the distal radius, followed by the proximal and distal tibia.

The presence of at least 10 GRLs in the long bones of children between 0 and 24 months of age is highly specific for maltreatment in the absence of another identified stressor. It is important to note that these findings most likely represent previous episodes of unidentified maltreatment, rather than the incident that resulted in their identification as a victim. This was the first study to identify the distal radius as the bone site most likely to have GRLs in physically abused children. Xrays with multiple growth recovery lines should raise concern for child abuse and neglect. week follow-up survey. Our routine skeletal survey includes AP and lateral radiographs of the long bones. The surveys were reviewed independently by two pediatric radiologists. Likert score from 1 to 5 (1=no CML, 5=definite CML) was used. Only cases with a Likert score of 4 or 5 by both radiologists were selected.Demographic, clinical, and imaging findings were recorded. One radiologist reviewed all CMLs at initial and 2 week follow-up skeletal survey for presence of the following signs: corner fracture, thin bucket handle fracture (BHF), thick BHF, BHF with endochondral bone filling the gap (BHFG), subphyseal lucency (SPL), increased metaphyseal density (IMD), deformed corner (DC), and subperiosteal bone formation (SPBF).We hypothesized that findings seen only on initial surveys represent acute phase signs. Any signs seen on 2 week follow-up represent healing phase.

The study group included 25 children (12 females) with age range 1-12 months (mean 3 months Purpose or Case Report: In physically abused children, fractures often go undetected; in such cases, time since injury (TSI) estimation may be essential for identification of abuse as well as prevention of continued violence. The present study was designed to systematically review empirical studies of radiographic methods to assess TSI of healing fractures in pediatric patients. These methods were evaluated for their utility in clinical as well as forensic settings, with particular concern for child protection.

A systematic literature search was performed of EBSCO, Embase, MEDLINE (PubMed), and Web of Science for scientifically-based radiographic methods to assess TSI published from the earliest available through August 6, 2018. After screening 4,549 articles, this search identified eleven empirical studies of pediatric fracture healing that met study inclusion criteria. Results: Of eleven articles, seven were based on samples of patients less than one year old while four articles combined patients of varying ages. Seven articles based their timelines on pooled fracture locations, of which six included fractures of the upper and lower limb; only two examined differences in healing based on fracture location. Two articles focused on abuserelated fractures while many articles failed to exclude abuserelated fractures. Inconsistencies in fracture healing variables exist across articles, which limits many direct associations. When comparisons can be made, healing timelines vary between articles, potentially in part due to the retrospective nature of the research. Though several articles mentioned that the methods and associated timelines could be applied to fracture locations other than those used in their study, few validation studies exist. In addition, the appropriateness of applying fracture healing timelines derived from accidental fractures to abuse-related fractures has not yet been explored. Conclusions: TSI methods applied to fracture locations and patients with ages other than those the method was developed from could provide inaccurate estimates of fracture healing.

Research into patient age, fracture location, and abuse status on fracture healing must be further examined before new methods are developed. It is suggested that methods derived from pooled patient age populations as well as pooled fracture locations (especially those derived from combined upper and lower limb fractures) be used with caution until otherwise substantiated for broader use. Animals were evaluated at 1 week and 6 weeks post implantation with CA, IVUS, black blood TSE MRI, contrast enhanced 3D radial MR angiography (MRA), MR 2D and 3D flow velocity mapping, and delayed enhancement imaging for fibrosis. In each subject, native IVC before and after the graft, proximal and distal graft anastomotic sites, and the mid graft were analyzed on all 3 modalities, and compared between weeks 1 and 6. Patterns of luminal distortion were assessed, and characterized using flow velocity changes, turbulence, energy loss, and tissue response as demonstrated on dynamic early contrast enhancement (DCE), and delayed enhancement (DE). Results: MRA yielded excellent morphometric definition of the native and grafted vasculature for luminal caliber and distortion when compared to angiography. There was good correlation between measurements of graft thickness obtained by IVUS and TSE. Some degree of stenosis was noted in all subjects with anastomotic and whole graft involvement, while there were no cases of rupture, aneurysm formation or infection. Unique information obtained from MRI included area deformation of the TEVG throughout the cardiac cycle, flow alterations and collateralization related to stenosis, estimates of wall shear stiffness and power loss, and estimate of host response at 1 and 6 weeks based on DCE and DE of the graft wall. Conclusions: MRI/MRA provides comprehensive assessment of TEVG by providing information on luminal architecture, wall thickness, graft integrity, flow perturbations and energy loss, which will be used to develop computational models to predict tissue engineered graft remodeling and stenosis. Purpose or Case Report: Dynamic contrast enhanced MR lymphangiography (DCMRL) is a well described technique of imaging the central lymphatics following intranodal contrast injection. Intrahepatic dynamic contrast enhanced MR lymphangiography (IH-DCMRL) involves ultrasound guided injection of a gadolinium contrast agent into the intrahepatic lymphatic ducts followed by MRI of the chest and abdomen with dynamic time resolved and delayed imaging. We aim to describe the MRI findings of IH-DCMRL.

Imaging of all patients less than 20 years of age who underwent an IH-DCMRL over 6 months was retrospectively and independently reviewed by 2 blinded pediatric radiologists who subspecialize in lymphatic imaging. Initially, success of intrahepatic lymphatic access and injection was evaluated. Imaging finds such as pericholecystic enhancement, mesenteric and retroperitoneal reflux, and visualization of the thoracic duct (TD) was assessed. Presence or absence of abnormal lymphatic perfusion to the peritoneal cavity, bowel, and lung and hila was determined. Images were also evaluated for hepatic vein contamination and peritoneal contamination. Disagreements between the 2 readers were solved by consensus. A chart review was performed for demographics and history. Results: A total of 21 patients (10 male) with a mean age of 7.4 years (range 4 months to 16.9 years, interquartile range 13.1 years) were identified. Clinical indications included pleural effusion (n=11), ascites (n= 7), protein losing enteropathy (PLE,n=6), anasarca (n=2), and intestinal lymphangiectasia (n=1). Two (10%) IH-DCMRLs were technically unsuccessful due to needle displacement. Pericholecystic enhancement (n=7) and reflux into the mesenteric (n=17) and retroperitoneal lymphatics (n=17) was present. Abnormal lymphatic perfusion to the peritoneal cavity (n=7), bowel (n=7), and lung (n=11) was seen. Readers were in agreement on the presence or absence of lymphatic perfusion to the peritoneal cavity in all cases, in 18/19 cases of perfusion to the bowel, and in 17/19 cases of perfusion to the lung and hila. The TD was visualized in 15/19 patients (79%). Hepatic vein contamination was common (n=14). Peritoneal contamination occurred in 3 patients, 2 of which were unsuccessful IH-DCMRLs. Conclusions: IH-DCMRL is a promising technique to evaluate the liver and central lymphatics which may be advantageous when lymphatic imaging via an intranodal approach is ineffective in demonstrating these lymphatic pathways. The inclusion criteria was normal cardiac anatomy, as well as normal preload and afterload. Epi and endocardial contours (EC) were drawn on the end-diastolic short-axis bSSFP images of the LV from mitral valve annulus to apex. These contours were used for automatic extraction of the trabecular edges (T). Following quantitative indices were computed using automated tool: 1) NC/C length ratio (LR) perpendicular to EC at each slice; 2) MR = percent NC of (NC+C) for entire LV; and 4) perimetric ratio (PR = length of T / perimeter of EC) for each slice. Global quantitative indices were derived as: 1) maximum of LR ( Purpose or Case Report: While Gadolinium-based contrastenhanced MR Angiography (Gd-MRA) is a robust technique for evaluation of the vascular system, there are inherent limitations, such as the contraindication in renal failure and concern of intracranial Gadolinium deposition. This study exams the feasibility of flow-independent relaxation-enhanced MR Angiography without contrast and triggering (REACT) in various clinical indications and compares the sequence to corresponding Gd-MRA.

The REACT sequence is a combination of a two-point 3D chemical-shift water-fat separated mDIXON TFE pulse sequence, a non-volume selective adiabatic inversion pulse, and a four-refocusing-adiabatic-pulse T2-prep module, which suppresses signal from static tissues and enhances the long T1 and T2 native signal of unenhanced blood, thereby providing optimal vessel-to-background signal contrast.We retrospectively reviewed all cases where both the REACT sequence and Gd-MRA were available for comparison between June 2018 and October 2018. Imaging was performed on a 3 Tesla platform (Philips Ingenia, software R5.3) or a 1.5 Tesla platform (Philips Achieva, software R5.3). Two pediatric body radiologists in consensus qualitatively compared results from REACT with Gd-MRA to evaluate whether it was diagnostically useful either as a complement or potential replacement. Results: We retrospectively reviewed a total 5 cases with both the REACT sequence and Gd-MRA. The clinical indications included pre-operative evaluation for vascular access, thoracic outlet syndrome, May-Thurner syndrome, deep venous thrombosis, and complex venolymphatic malformation. The REACT sequence was found to be equivalent at obtaining the diagnosis in all 5 cases. However, in 4 of the 5 cases, Gd-MRA provided improved contrast resolution and signal-to-noise compared with the REACT sequence. Only in the case of thoracic outlet syndrome were the images found to be noninferior. Conclusions: Our preliminary experience with REACT in children suggests that the sequence is robust and capable of providing diagnostic angiograms in a variety of clinical indications. One advantage of REACT is that it can be repeated as needed and is not constrained by timing of a contrast bolus. However, the major disadvantage of REACT is its lower contrast resolution and signal-to-noise ratio compared with Gd-MRA. Future studies with larger and more diverse sample sizes are needed to better validate whether REACT can obviate the need for Gd-MRA in more generalizable circumstances. Purpose or Case Report: Time-resolved, 3D phase-contrast (4D flow) MRI enables comprehensive cardiac evaluation, but long acquisition times and motion corruption limit pediatric abdominal clinical use. Here, we develop and integrate a timeefficient non-Cartesian 3D cones trajectory into 4D flow to reduce scan time while also increasing robustness to respiratory and bowel motion artifacts. We then assess whether the resulting non-Cartesian 4D flow is able to achieve high-quality, full-coverage cardiorenal evaluation in under 10 minutes.

We design a flow-encoded, goldenangle re-ordered 3D cones pulse sequence based on RF-spoiled gradient recalled echo (SPGR). Cones gradient waveforms are iteratively designed on-the-fly to enable arbitrary field of view and resolution prescription. Cardiac-resolved images are reconstructed using a combined parallel imaging and compressed sensing algorithm (l1-ESPIRiT). To further improve motion-robustness, respiratory signals are estimated from each cone readout, and then used to suppress motion during reconstruction.With informed consent and IRB approval, 2 pediatric subjects (3-years-old and 7-years-old) referred for contrast-enhanced abdominal MRI were scanned using Cartesian 4D flow and cones 4D flow sequences on a 3T scanner (MR750, GE Healthcare) with a 32-channel cardiac coil. Cartesian and cones 4D flow scan parameters include spatial resolution: 1.0x1.0x1.5 mm 3 , 8 cardiac phases, Venc: 100-150 cm/s, and scan durations: 5-7 minutes. All data was acquired with subjects freely breathing. Results: Cones 4D flow reconstructions are feasible and show excellent delineation of renal vasculature in data acquired just under 5 minutes. By nature of the cones sampling trajectory, respiratory and bowel motion artifacts appear noise-like and incoherent. Similar image quality is observed after retrospectively undersampling data to simulate a 3-minute acquisition. Further, both aortic and renal arterial flow rates are conserved in the retrospectively undersampled acquisition. In a 7-minute acquisition, where the Cartesian and cones scan times are held constant, cones 4D flow provides superior image quality, but underestimates peak aortic flow by 26%. This difference is attributed to larger eddy current errors in the cones sequence, although, these can be corrected with more robust physical models.

We introduce the first 4D flow sequence with a 3D cones sampling trajectory. Cones can highly accelerate 4D flow acquisitions and reduce both flow and motion artifacts. Purpose or Case Report: Multidimensional imaging approaches reduce total examination time by using a single acquisition to reconstruct several images contrast states, typically requiring several independent scans. In children, this scan reduction increases patient comfort and compliance while reducing the need for anesthesia. In this work, we introduce a novel multidimensional imaging approach that utilizes a rosette, or flower-like, k-space trajectory entitled Contiguous Rosette Echoes iN Single Highly Accelerated Acquisition (CRENSHAA). Using CRENSHAA, we demonstrate feasibility for self-gated, motion resolved T2* mapping for simultaneous cine imaging and iron quantification.

TheoryRosette trajectories employ oscillating gradient fields at two distinct frequencies to produce flower-like k-space traversal patterns first described by Doug Noll [1]. We extend this earlier work by using a novel rosette parameterization: If N is odd, q = { x | x = (2L-1)/N, L = (N+1+2m)/2, m Î Z + } \ Z + If N is even, q = { x | x = (2L)/N, L = (N-2+4m)/2, m Î Z + } \ Z + where N represents the number of petals, L the number of layers, m is an incrementing variable and Z+ is the set of all positive integers. MethodsAll scans were performed on a GE 750W with a 20 channel cardiac array. Ex Vivo: T2* phantoms containing distilled water, MnCl2 and 3% carrageenan in 50ml falcon tubes were constructed for T2* validation. A CRENSHAA sequence with q=2.2, FA= 15 degrees and TR 34 ms produced echoes at 1.1, 4.1, 6.7, 9.3 and 11.9ms. 350 repetitions were acquired, each successively incremented by a rotation angle of 137.5° Exponential fitting was performed in MATLAB. Images were compared to a gold standard GRE multiecho sequence. In Vivo: A 31 y/o male volunteer was imaged using a CRENSHAA sequence. Selfgated, k-space magnitude CINE images were reconstructed at 30 phases with a 5 phase sliding window. All studies were IRB approved and conducted with informed consent. Purpose or Case Report: Cardiac CINE is a common MRI technique for cardiac evaluation, but requires multiple breathholds to acquire high quality images without respiratory motion artifacts. This in conjunction with rapid heart rate makes pediatric CINE imaging challenging. Neural networks can leverage previous exam data to learn how to reconstruct highly accelerated MRI data. Here we extend this to cardiac imaging, and present a deep learning-based reconstruction technique trained on fully-sampled healthy volunteer data. We show that our network outperforms combined parallel imaging and compressed sensing (PICS) reconstruction methods on 10x accelerated 2D CINE data with respect to common image quality metrics.

The proposed unrolled network iteratively applies 3D convolutional neural networks and data consistency updates to under-sampled input data. Data consistency layers enforce consistency with input k-space samples, and help the network generalize to data acquired from unhealthy patients. The network is trained to output fullysampled complex images, preserving both magnitude and phase.With IRB approval, fully sampled 2D bSSFP cardiac CINE datasets were acquired from 12 healthy volunteers at different cardiac views and slice locations on 1.5T and 3.0T GE scanners. For training, ten volunteer datasets are split slice-byslice into 155 slices, and then further augmented by random flipping, cropping, and variable-density undersampling (R=8-10). For evaluation, the remaining two volunteer datasets are retrospectively undersampled to simulate 10-fold acceleration with 25% partial echo. The same evaluation datasets are reconstructed using the proposed method and a PICS algorithm (l1-ESPIRiT). Reconstruction quality is evaluated with respect to peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM).

The proposed method outperforms PICS with respect to both PSNR and SSIM metrics. The improvement of the proposed method over PICS is most apparent inside the heart with less spatiotemporal blurring of myocardium and papillary muscles in deep reconstructions. Deep reconstructions also depict systolic heart motion much more naturally, whereas PICS is susceptible to temporal staircasing artifacts due to strong regularization.

Preliminary results suggest that the proposed method can reconstruct 2D cardiac CINE data more accurately and robustly than compressed sensing. This technique can improve image quality for single breath-hold, and in the future, free-breathing scans.

The utility of the ASL sequence in parenchymal injury of the brain in abusive head trauma (AHT). 

Our study demonstrated that in cases of suspected abusive head trauma, evaluation of the ASL sequence in addition to other MRI sequences provided additional information regarding parenchymal injury as compared to routine MRI sequences alone. We propose that by adding the ASL sequence, in the setting of suspected AHT, to the evaluation brain parenchymal injury would add value by improving the detection of injury, assess long term prognosis, and may also help understand the mechanism of injury. Purpose or Case Report: Congenital Zika syndrome (CZS) is unique to fetuses infected with Zika virus (ZIKV) before birth and presents with distinct pattern of birth defects including characteristic brain abnormalities. The neurological manifestations in these infants with CZS have emerged lately and are well documented. However, the consequences of ZIKV exposure to fetuses of pregnant mothers and the neonates born to them were not thoroughly examined. Despite these ZIKVexposed neonates demonstrating normal head circumference at birth [1], they may have subtle brain abnormalities and neurodevelopmental deficits that are possibly associated with congenital ZIKV [2,3]. Since these ZIKV-exposed infants represent a larger population as compared to those with CZS, it becomes all the more important to examine the brains of this group. The purpose of this prospective study is to compare the brain MRI volume measures of otherwise normal appearing ZIKV-exposed infants to matched controls, with neurodevelopmental correlates.

Volumetric MRI data of ZIKV-exposed infants (n = 9) and normal controls (n = 5), acquired from 1 to 6-months gestational age were scanned without sedation or contrast on a 3T MRI scanner. The MR protocol consists of T2, diffusion kurtosis imaging (DKI), susceptibility weighted imaging, and MR Spectroscopy (MRS) sequences. The images were hand contoured using Multitracer software to quantify the brain parenchymal and CSF volumes. The volumteric data of the above groups was compared, and the volume data of the ZIKV exposed group was correlated with outcomes of the NICU Network Neurobehavioral Scale administered at full-term equivalent gestation and the Bayley-III assessment performed at 6 months of age using the Student's T-test and Pearson's correlation test.

The findings in the ZIKV-exposed neonates include a) significantly decreased brain parenchymal and supratentorial volumes and increased CSF volume as compared to matched controls, and b) a positive association between brain parenchymal volume and Bayley-III Language and Cognitive Composite Scores. Conclusions: Our preliminary results demonstrate some of the neonates exposed to ZIKV show Zika-associated brain and neurobehavioral abnormalities. Further evaluation of the DKI and MRS sequences will allow for additional conclusions regarding subtle changes in tissue microstructure and metabolic activity, respectively. Longitudinal collection of brain imaging and behavioral data will further refine the clinical impact of Zika disease. Purpose or Case Report: Surgery is an important option in the management of medically refractory epilepsy. Yet even in ideal surgical cohorts, a significant fraction of patients (typically 35-45%) do not achieve seizure freedom. This failure reflects in large part that pediatric focal epilepsy is a disorder of widespread cerebral cortical networks; some of these network alterations are epileptogenic and can drive recurrent seizures after resection. The goal of this study, therefore, was to define the relationship between global network architecture and seizure-free outcome in children selected for surgical management of focal epilepsy.

This is a retrospective, IRB approved study. Patients were identified who underwent surgical resection for seizure management and had 3Tesla-MR before surgery, including rs-fMRI. Surgical outcomes were assessed using the Engel classification: 1. free of disabling seizures; 2. rare disabling seizures; 3. some improvement; 4. no improvement.

Resting state fMRI images were co-registered to a T1-weighted structural image, corrected for motion, and high-pass filtered (0.01 Hz). Networks for each child were defined using an anatomic parcellation technique with subdivision of whole brain gray matter into approximately 750 nodes. An undirected graph was constructed based on the pair-wise correlation of node BOLD time series. The following topological properties were calculated: clustering coefficient, modularity, path length and efficiency. A multivariate statistical learning technique was used to measure the independent contribution of each metric (adjusting for age, sedation during MR, gender, seizure duration and pathologic diagnosis) to surgical outcome. Results: Fifty-two patients met inclusion criteria (21 female, mean age 10.3 ± 5.42 years). Diagnoses included primarily focal cortical dysplasia (n=21) and mesial temporal sclerosis (n=15). Global efficiency, clustering coefficient, and pathologic diagnosis each made an independent contribution to prediction of Engel outcome by the learning algorithm. Prediction of Engel class 1 outcome was driven by global efficiency; Engel class 4 prediction by clustering coefficient.

We observed characteristic global network signatures of good/poor outcomes after epilepsy surgery. These findings support the potential clinical relevance of brain network metrics in children with refractory epilepsy. Purpose or Case Report: Dystonia is one of the most common paediatric movement disorders, and often the most difficult to manage. Deep brain stimulation (DBS) is a surgical treatment that modulates dysregulated motor circuits, and has been used in 321 medically refractory cases, with mixed outcomes. Factors preventing maximal benefit may include: heterogeneous disease aetiologies, suboptimal lead placement, and difficulty in programming. Surgical insertion currently relies on adultderived techniques and anatomic targeting. During programming, the clinician chooses which electrical settings, e.g. voltage, provide the greatest clinical benefit to an individual patient. Choosing these individualised parameters can be a tedious and often imprecise trial-and-error process. It requires multiple clinic visits, and reliance on difficult clinical observations. With appropriate targeting and stimulation programming, DBS can produce striking clinical benefits. We hypothesized that the imaging of our DBS-implanted paediatric dystonia patients could inform us on which neural networks and tracts should be modulated, resulting in improved clinical outcomes.

We conducted a retrospective analysis of 11 paediatric dystonia patients who underwent Globus Pallidus internus (GPi) DBS insertion at our institution. DBS electrodes were localized and transformed into a normative brain space. Volume of tissue activated (VTA) was estimated and weighted according to their associated clinical outcomes. Normative data were then used to identify functional and structural networks associated with optimal clinical benefits.

The majority of our patients demonstrated clinical improvement. Using standardized neurosurgical coordinates, the optimal area of stimulation was located at 22, -8, -2, within the posterior GPi. This anatomical locale resulted in the best clinical benefit and lowest side effect profile, as determined clinically. The most closely associated neural network areas were the bilateral basal ganglia, cerebellum, and prefrontal cortex, all of known importance in motor circuitry. White matter tracts associated with the greatest clinical improvement were also involved in motor circuitry. Conclusions: Paediatric dystonia patients responded better to DBS treatment engaging specific networks and tracts of the motor circuit. These findings could be used to improve surgical technique and post-operative empiric electrical titration, thereby benefiting patients and their families through better care and clinical outcomes. Life-long mTORi therapy with extended imaging monitoring may be required to maintain therapeutic benefit. In the setting of increasing concerns regarding gadolinium retention in the body, we aimed to preliminarily assess utilization of non-GBCMenhanced MRI sequences for identification of clinically meaningful changes in SEGA volume.

Following IRB approval, the medical records and imaging for an internal database of TSC patients seen at our tertiary pediatric referral center in the last 2 years were reviewed. Patients with SEGAs demonstrating "clinically meaningful growth" (resulting in SEGA diagnosis or offer of tumor directed intervention) on serial MRI exams were identified. Paired exams were reviewed by a subspecialty board certified pediatric neuroradiologist to determine changes in tumor volume first on non-GBCM-enhanced sequences (T2W, T1W), and subsequently on GBCM-enhanced T1W sequences. Wilcoxon rank sum tests were performed in R. Results: Of 68 patients in the TSC database, 9 (13%) had SEGAs. 6 of these patients had 7 SEGAs that demonstrated clinically meaningful growth on serial MRI exams. For these exam pairs, the distribution in measured changes in tumor volume was not significantly different when assessed with non-GBCM-enhanced sequences versus GBCM-enhanced sequences (p= 0.96), or between non-GBCM-enhanced sequence assessments (p= 0.88). Changes in tumor volume ranged from 148% to 115,050% increase. Conclusions: TSC patients often undergo a large number of GBCM-enhanced MRI exams. In our study, assessments of changes in SEGA volume on MRI were not significantly different between GBCM-enhanced and non-GBCM-enhanced sequences. Given the lack of clinical benefit, it is prudent to limit GBCM administration for routine MRI follow up of subependymal nodules and known SEGAs in patients with TSC. Purpose or Case Report: Pediatric patients with optic pathway gliomas (OPGs) undergo a large number of follow-up MRI brain exams with gadolinium based contrast media (GBCM). Gadolinium retention in children has motivated parsimonious use of GBCM. We previously determined that increased OPG size was the only isolated finding on MRI that motivated changes in tumor directed therapies. In a pilot case series, change in tumor size was readily identifiable on non-GBCM enhanced MRI sequences. Our purpose was to investigate this result in a blinded, non-inferiority, multi-reader-multi-case study.

Following IRB approval, power calculations were performed with pilot data. The primary endpoint was intra-reader agreement for ≥25% increase in greatest axial diameter cross-product measurement. A denominator OPG population was derived from a regional cancer registry at our tertiary pediatric referral center. Consensus tumor board notes were reviewed for each patient, with attention to: (1) isolated OPG status, (2) change or (3) stability in OPG size between two specific MRI exams. All instances of criteria 1 & 2 were incorporated in the case set. Cases meeting criteria 1 & 3 were incorporated in order of most recent clinical follow up, maintaining a 50% ratio of NF1:non-NF1 exams. Exams were standardized with non-GBCMenhanced and GBCM-enhanced versions. 7 pediatric radiologists were assigned to a randomized case panel for 3 blinded sessions, spaced by at least 1 week. The first session excluded GBCM-enhanced sequences, the others did not. Interim statistical analysis was performed in R for 4 readers. Results: Power calculations revealed that a ≥80% probability of rejecting, at the α= 0.05 level, the null hypothesis of -12% (noninferiority margin) difference in intra-reader assessments beyond background variability could be achieved with an "n" of 60 exam pairs, each interpreted by at least 4 readers. Exam pairs were compiled from 42 patients with isolated OPG (19 with NF1), from a population of 104 patients with OPG. There were 8 "size change" exam pairs from 7 NF1 patients and 20 "size change" pairs from 18 non-NF1 patients. Tumors encompassed a diverse anatomic spectrum. Interim analysis demonstrated an overall -3% difference (95% CI: -9.9, 3.1%) in intra-reader agreement when using a non-GBCM-enhanced protocol.

Conclusions: Interim analysis suggests that a non-GBCMenhanced protocol is non-inferior to a GBCM-enhanced protocol for assessment of change in size of isolated OPGs on follow-up MRI exams. Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Background: Primary brain sarcomas are rare, with an incidence reported ranging from 0.1 to 4.3% of all cerebral tumors and with a high mortality. There is a paucity of literature depicting radiologic features in children. Objective: Describe the most frequent MRI findings of pediatric intracranial sarcomas.

We retrospectively analyzed the data of 12 consecutive pediatric patients (7 boys and 5 girls; mean age: 8 years) referred to two national paediatric centres who underwent MRI imaging within a 48-month period.

The most frequent MR feature were the presence of hemorraghe and necrosis, both in 92% (n=11); diffusion restriction and presence of cystic component, both in 83% (n= 10); regular borders and homogenous intense enhancement, in 75% (n= 9) ; and finally 67% (n=8) of tumors were supratentorial and showed extension to the meninges. Perfusion was performed in 7 patients, all of them with increased perfusion. 8 patients had spectroscopy, 50% (n= 4) of this cases showed increased lactate, lipids and choline and decreased NAA.

In the setting of pediatric brain tumor with regular edges, diffusion restriction, hemorrhage, cystic component, homogenous intense enhancement and increased perfusion, we must consider primary intracranial sarcoma in the differential diagnosis.In the setting of pediatric brain tumor with regular edges, diffusion restriction, hemorrhage, cystic component, homogenous intense enhancement and increased perfusion, we must consider primary intracranial sarcoma in the differential diagnosis. Purpose or Case Report: Volumetric fast spin echo (FSE) of the knee using T2 shuffle (T2Sh) technique has previously been described as being comparable to traditional 2D imaging. T2Sh has the added advantage of being a rapid single-scan 4D multiplane reformattable sequence for pediatric knee examinations. The purpose of this study is to investigate the feasibility and effectiveness of a targeted rapid pediatric knee MRI exam after introduction into clinical practice, with the goal of reducing cost and enabling same-day MRI access.

In an institutional review board approved study with informed consent/assent, we implemented a targeted pediatric knee MRI exam on three 3T scanners for assessing pediatric knee pain. The 10-minute protocol was based on T2 Shuffling, a four-dimensional acquisition that permits volumetric reconstruction of images with variable T2 contrast, and a single T1 2D FSE sequence. To enable a clinically feasible image reconstruction time, a distributed, compressed sensing-based iterative reconstruction was implemented on a local four-node high-performance compute cluster and integrated into the clinical scanner and PACS workflow. Pediatric patients were sub-selected for the exam based on insurance plan and clinical indication. Over a two-year period, 47 subjects were recruited for the study and 49 MRIs were ordered. Date and time information was recorded for MRI referral, registration, and completion. Descriptive statistics were also analyzed. Image quality was also assessed from 0 (nondiagnostic) to 5 (best anatomy delineation) by two radiologists, consensus was subsequently reached. Purpose or Case Report: Juvenile dermatomyositis (JDM) causes diffuse vasculitis and proximal muscle inflammation. Searching for highly sensitive non-invasive methods for early diagnosis and treatment follow-up is key in this disease. We investigated blood-oxygen level-dependent (BOLD) MRI to detect physiologic changes at different levels of severity of JDM as compared to age-matched healthy control subjects.

Six JDM patients of 11-15 years (12.9 ± 1.6 years) and 6 age and sex-matched healthy controls (13.29 ± 1.50 years; 6 males and 6 females) completed the study. Participants performed 3 cycles of one-minute up-down exercise on a MRI-compatible cycle ergometer at 65 % of maximum workload, with 2 minutes of rest between each bout. BOLD MRI images were acquired immediately following exercise at 3T. Parameters of 10-min BOLD EPI gradient echo sequence were: TR/TE = 250/40 ms, flip angle = 90 0 , FOV = 25 mm, matrix = 64x64 and slice thickness/gap = 10/0 mm. In addition, Childhood myositis assessment scale (CMAS) was collected as an indicator of disease progression. BOLD signal response in the vastus medialis (VM) and intermedius (VI) muscles during recovery were fitted to a sigmoid model with β, half-time point for recovery of intensity curve; α, response time and wash-out of deoxyhemoglobin; and κ, range of recovery from baseline. in JDM compared to healthy subjects following recovery from exercise.

Results: A comparison of BOLD signals between control and JDM subjects in VM and VI demonstrated a significant difference for the first half-time point for β of VM (p = 0.04). In addition, when adjusted for the CMAS score, the second recovery β for VI and the first for VM were significant (p = 0.03 and 0.02). However, all other variables showed no significant difference after adjusting for CMAS.

Our BOLD MRI sigmoid model at 3T allowed us to differentiate vascular and oxygenation changes in the quadriceps muscle between JDM and control cohorts in recovery re-oxygenation after exercise. Therefore, BOLD MRI may be useful to characterize early disease and monitor effects of therapy or exercise as adjuncts to conventional laboratory and clinical assessments of JDM. Purpose or Case Report: Magnetic resonance imaging-guided high-intensity focused ultrasound (MR-HIFU) is a novel technology that integrates magnetic resonance imaging with therapeutic ultrasound. Although clinical experience in pediatrics is relatively small, the advantages of a completely noninvasive, precise, and radiation-free tumor therapy is especially attractive to growing children. Pediatric radiologists should be familiar with clinical applications of MR-HIFU, including pre-and post-procedure imaging protocols, treatment imaging, and post-treatment imaging assessment. To this end, we review applications of MR-HIFU in children and present cases from our institution to highlight the post-treatment MRimaging findings and evolution of treated lesions over time.

Under IRB approved clinical protocols, 15 patients (age 7-21 years) with symptomatic benign (osteoid osteoma), locally aggressive (Desmoid tumor) and refractory malignant (Sarcoma) tumors have undergone MR-HIFU therapy at our institution since 2017. Pre-treatment diagnostic MRI scans, treatment imaging, and post-treatment follow-up MRI scans were reviewed. In addition to standard musculoskeletal MR sequences, volumetric 3D T2 weighted sequences as well as pre-and post-contrast axial fat-saturated T1 weighted sequences with subtraction imaging were obtained. Imaging findings were correlated with clinical history and follow-up course.

Results: We will present MRI findings from patients treated with MR-HIFU. Preprocedural, intraoperative, and postprocedural imaging findings will be discussed, and sample MRI protocols will be presented, detailing the purpose of essential sequences. Conclusions: Our preliminary institutional experience with MR-HIFU has been educational for our pediatric radiology and oncology teams. There are currently ongoing studies for MR-HIFU therapy for a treatment of a variety of tumors as well as for enhanced local drug delivery. As use of this novel therapy expands in the pediatric population, it is important for diagnostic radiologists to have a basic understanding of the nature of the procedure as well as what to expect on imaging studies. Furthermore, close collaboration between radiologists and oncologists will be essential to determine the relationship between imaging findings and clinical response. (70%) were followed a minimum of 1 year post-reduction (12-45 months) and assessed for presence of AVN. Three hips developed AVN (21%; 95% CI = 6 to 51%). Based on ROC analysis, diagnostic utility was nearly the same for CEUS and CEMRI ( = 0.74; 95% CI = 0.35 to 1.00 and AUCMRI = 0.77; 95% CI = 0.39 to 1.00, respectively). AVN was diagnosed at follow-up where decreased enhancement was noted on CEUS in 71% (67% sensitivity; 73% specificity) and on CEMRI in 77% (67% sensitivity; 80% specificity). This is similar to previous CEMRI reports after closed reduction.

Conclusions: CEUS appears to be a viable method of assessing infant hip perfusion with overall good correlation with postoperative CEMRI. Further study is required to determine whether CEUS will provide adequate real-time feedback to aid in successful reduction while reducing the incidence of postoperative AVN. Purpose or Case Report: To demonstrate the feasibility and accuracy of automatically extracted sonographic skin-to-livercapsule distance (SCD) measurement as a biomarker for fatty liver in the pediatric population.

This institutional review board approved single-institution retrospective study evaluated consecutive pediatric patients, ages 2 to 18, who underwent abdominal ultrasound from February 2018 to September 2018. Diagnosis of fatty liver, obtained through the electronic medical record, was recorded on the basis of clinical notes, prior imaging tests, and liver biopsy pathology when available. Body mass indices (BMI) and normalized Z-scores using CDC's growth charts, were also extracted. Duplicate patients and those with absent clinical information or imaging were excluded. Using manually measured SCD as a reference standard, an image processing algorithm was developed to automatically extract the SCD measurement from manually selected B-mode images (one image per patient). Image processing steps included: 1) Region of Interest (ROI) placement, 2) image denoising (Gaussian filter), 3) contrast enhancement, 4) canny edge detection, 5) connected components detection using the binary image derived from step 4, 6) skin and Glisson capsule border detection, and 7) SCD calculation (distance between the detected skin and capsule). Receiver operating characteristic (ROC) curve analysis was performed to assess the diagnostic performance of BMI Z-scores and the SCD measurements for fatty liver. Results: 174 patients, including 26 with hepatic steatosis, were analyzed (mean age 125.6±54.1 months, 74 males, 100 females). The root-mean-squared error (RMSE) between manually and automatically extracted SCD measurements was 5.06 mm. The area under the ROC curve (AUROC) for prediction of fatty liver was calculated to be 0.905 (95%CI: 0.856-0.954) for the BMI Z scores, 0.889 (95%CI: 0.818-0.961) for the manual SCD measurements, and 0.770 (95%CI: 0.668-0.873) for the algorithm-measured SCD.

Conclusions: In addition to BMI, sonographic SCD can be considered as a biomarker of hepatic steatosis. Measurement of SCD can be automated with acceptable accuracy. The benefits of using sonographic SCD rather than liver echogenicity as a biomarker are threefold: 1) it is faster, 2) can be performed at the point of care, and 3) reduces operator dependence compared with interpretation of B-mode ultrasound liver images. Purpose or Case Report: To compare biexponential analysis of T2* relaxation of the liver in children with iron deposition disease to the more commonly used monoexponential model and establish its relation with different levels of iron overload in children.

All MRI studies performed for determination of liver iron concentration (LIC) performed in our institution between 2007 and 2017 that included T2* sequences were included. LIC was calculated using Monoexponential T2* (MonoExp) and Biexponential T2* (BiExp) models using both commercially available and an in-house developed software. The calculations were based on ROIs including the entire liver (ROI1), the periphery of the liver (excluding the major vessels) (ROI2) and the spleen (ROIcontrol). The LIC results were classified as normal (< 3.2 mg/dL), mild (3.2 -7 mg/dL), moderate (7 -15 mg/dL), and severe (> 15 mg/dL) as per commonly used treatment guidelines. Residuals and residuals ratios were calculated for each model (MonoExp and BiExp) in order to identify the best fitting model on each LIC category. A MonoE/BiExp residual ratio > 1.5 was considered to have a biexponential trend. Non-parametric tests and correlations were used, p < 0.05 was considered significant. Results: 182 patients (110 boys, 71 girls) with a mean age of 12.3 years (range or SD) were included in the analysis. For both ROI1 and ROI 2, there was a near perfect agreement between MonoExp using the commercial and the in-house software (r = 0.98, p < 0.001, and r = 0.99, p < 0.001, respectively). According to the MonoExp/BiExp residual ratio, the BiExp model fits better in patients with a LIC > 15 mg/dL (ROI1: ratio = 11.5, p < 0.001 and ROI2: ratio = 2.3, p < 0.001). For those with a biexponential best fit (n = 38), the proportion of slow (second) component was 0.06 (range 0.01 -0.11). In the internal control ROI (spleen), the MonoExp/BiExp ratio was not significantly different among different LIC categories (p = 0.46). Conclusions: Biexponential T2* relaxation model for LIC estimation fits better than the monoexponential model in patients with severe iron overload (LIC > 15 mg/dL) and can be used to more accurately determine LIC in these patients. The proportion of slow (second) component of the model might be related to the intra-cellular iron in the liver. Many studies have demonstrated that liver stiffness directly correlates with liver fibrosis. Several studies of liver fibrosis in post mortem autopsies have proven that liver fibrosis is a heterogeneous process. There is an abundance of scientific literature evaluating global liver stiffness. Few studies have assessed regional liver stiffness variations in children using the eight Couinaud liver segments.

This retrospective IRB approved study involved 170 children who underwent MRE at 3T on a GE platform using a commercial 2D SE-EPI pulse sequence. Regional liver stiffness was measured at the locations of the Couinaud segments by a trained analyst. In each patient, the segment exhibiting maximum stiffness was recorded. The coefficient of variation (CoV=standard deviation/mean) across the eight segments was also computed. A higher CoV signifies greater stiffness heterogeneity. Medical history was obtained from patient records and the cohort was stratified into five major disease groups, including: Nonalcoholic Fatty Liver Disease(NAFLD) (N=119), autoimmune disease (N=22), heart and cardiovascular conditions (N=4), genetic disorders (N=12) and "others", which included cystic fibrosis, portal vein thrombosis, ductal plate malformations and liver tumors (N=13). Statistical analysis was performed using R software. Results: A wide range of regional variations in liver stiffness were observed (CoV=17.2 ± 7.3%, range =3.1-79.2%) across the cohort. Of the 170 patients, segments II and VII exhibited the highest stiffness, each in 29 patients. Patients with NAFLD displayed a significantly higher CoV (17.17±5.5%) compared to patients with autoimmune diseases (14.65±4.2%, p=0.019). No other significant differences in stiffness heterogeneity were observed.

We observed significant variations in liver stiffness across the eight Couinaud segments in 170 pediatric patients, with segments II and VII frequently exhibiting maximum stiffness. There were notable variations in stiffness heterogeneity between patients with NAFLD and autoimmune disease. Our results suggest that underlying disease conditions can potentially influence the degree of regional stiffness heterogeneity. Measuring regional stiffness may also better guide targeted biopsies and interventions for the patient. Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

To know the usefulness of shear wave elastography (SWE) and dual energy computed tomography (DECT) for the diagnosis of hepatic veno-occlusive disease (VOD) in rabbit model.

Among six New Zealand white rabbits (3-4 kg, male), three rabbits had normal saline ingestion throughout 20 days and grouped as normal group. Another three rabbits had 6-thioguanine (6-TG, 5mg/kg/day) ingestion throughout 20 days and grouped as VOD group. Liver stiffness was measured using supersonic shear wave imaging on baseline, 3, 10, and 20th days. Liver perfusion was evaluated using DECT with fast kVp switching (80/140 kVp) on the same days. Three region-of-interests (ROIs) were drawn in liver parenchyma using virtual monochromatic images (VMI) of 55 keV and iodine maps. Morphologic changes of liver including periportal, gallbladder wall edema, or diameter change of hepatic veins were assessed on 20th day using CT. Final pathologic score of VOD was evaluated on the 21st day after sacrifice and compared between two groups using Mann-Whitney U test. Linear mixed model was used to know the differences of liver stiffness and perfusion parameters between two groups. Spearman correlation test was used to know the correlation between liver stiffness, perfusion parameters and pathologic scores. Results: Final pathologic scores were significantly higher in VOD group than in normal group (median 22, 2, p=0.046). There was no gross morphologic change in the livers in CT.

Liver stiffness values using SWE, Hounsfield unit (HU) values using VMI, and iodine concentration values using iodine map were significantly higher in VOD group compared with normal group on 10th and 20th days (all, p≤0.03). Compared to the baseline values, liver stiffness values and iodine concentrations became significantly higher in 20th day in VOD group (p=0.001, <0.001, respectively), while it were not different in normal group. Liver stiffness and iodine concentration showed positive correlation with final pathologic scores (all, r=0.754). Conclusions: Liver stiffness values from SWE and iodine concentration values from DECT were significantly increased in hepatic VOD model in rabbits. SWE and DECT could aid the diagnosis of hepatic VOD, while morphologic change was not apparent in the liver. Purpose or Case Report: Ultrasound shear wave elastography (SWE) is being increasingly used as a non-invasive means of measuring tissue stiffness which reflects pathologic processes including fibrosis, inflammation and congestion. Consensus guidelines suggest cut-off liver stiffness values for disease in adults but there are limited data on normal values for children. For the pancreas, understanding normal pancreas stiffness might be of value for future detection of diffuse pancreatic disease, including chronic pancreatitis. The purpose of this study was to define normal stiffness values measured in terms of ultrasound shear wave speed (SWS) for the liver and pancreas in the pediatric population and to assess for associations with patient sex, age and size.

In this prospective, IRB approved study, 121 healthy children with body mass index (BMI) <85 th percentile and without a history of liver disease underwent 2D SWE of the liver and pancreas with a Canon Aplio i800 ultrasound system using a i8CX1 probe. 10 liver stiffness measurements were obtained in the right hepatic lobe and 5 pancreas stiffness measurements were obtained in the pancreatic body or tail, all using a manually placed region of interest (1 cm in the liver, 0.5 cm in the pancreas). Descriptive statistics were utilized to summarize continuous data. T-tests and Pearson correlation were used to define relationships between SWS and predictor variables. Results: Mean age for the 121 children was 6.56±5.61 years (range: 0.02-17.81 years), mean BMI percentile was 39.7±26.8%, and 62 (51.2%) were female. The majority of children (87.6%, 106/121) were white, non-Hispanic. Mean median liver SWS for the population was 1.29±0.14 m/s. Mean median pancreas SWS for the population was 1.31±0.15 m/s. There were weak but statistically significant correlations between median liver SWS and age (r=0.29, p=0.001), height (r=0.32, p=0.00038), and weight (r=0.31, p=0.00045) but there were no significant correlations between pancreas SWS and any measure of subject size. There was no significant difference in liver or pancreas SWS based on sex and there was no significant correlation between liver and pancreas SWS.

We report normal liver and pancreas stiffness values for healthy children ages 0-17 years. Liver SWS >1.56 m/s and pancreas SWS >1.61 m/s likely reflect abnormal stiffening (>95 th percentile) that may be indicative of diffuse disease.

Purpose or Case Report: MRI is an imaging modality that provides a paradigm shift in the diagnosis of pediatric acute appendicitis concerning reduced scan time, lack of radiation and avoidance of intravenous contrast. Our study purpose was to compare devised MRI criteria for assessment of pediatric appendicitis with standard criteria.

A prospective, IRB approved study, using a non-sedated ultrafast non-enhanced MRI protocol consisting of coronal and axial fast spin echo T2 sequences without fat saturation and axial DWI. Expected maximum scan time was 10 minutes. MRI studies were reviewed independently by two radiologists and were graded as: positive, negative or equivocal for acute appendicitis using devised MRI criteria (based on retrospective review of confirmed positive cases) vs standard criteria (reports) as classifiers. Devised criteria for positive MRI were appendiceal diameter ≥10mm, or periappendiceal signal intensity (SI) on T2-WI, and presence of an appendicolith in the presence of other inflammatory signs. Absence of those criteria was graded as negative. Purpose or Case Report: The gender distribution of physicians has emerged as a significant policy issue in diagnostic radiology. While radiology has traditionally experienced female under-representation, pediatric radiology is a rare subspecialty in which men and women are represented similarly. Pediatric radiology is also unique in that approximately half of pediatric radiologists are affiliated with academic centers which renders the subspecialty rich with opportunities for teaching, leadership, and scholarly discovery. This study looks at current trends in male and female participation in recent educational, leadership, and research-based SPR-sponsored activities as a means to define the current state of gender representation within this subspecialty while developing baseline metrics and goals to promote a culture of inclusion.

A retrospective review of materials available to SPR members was performed. For the purposes of this study, male and female genders are defined by the most common associations of given names in the United States. In cases where the gender of the individual was in doubt, an Internet search for pronouns used in the physician's profile (he/him/she/her) was used to designate gender for the purposes of this study. Individuals in which determination could not be made were excluded. The 20 largest children's hospitals were defined by using Becker's Hospital Review. Results: Females are historically under-represented as past presidents of the SPR, but this trend has reversed in recent years. 50% of those scheduled to provide oral scientific presentations at the 2018 SPR Annual Meeting were female, but women represented only 36% of presenters in the Postgraduate Course. As of mid-2018, fewer than 5 of the 20 largest children's hospitals had female radiologists-in-chief. Regarding the 2018 meeting of the Society of Chiefs of Radiology at Children's Hospitals (SCORCH) which may host multiple representatives from a single hospital or include representatives acting in place of sitting chiefs, 37% were female.

Women remain under-represented in some SPRrelated roles. SCORCH attendance suggests that women are either more likely to lead smaller pediatric radiology departments or serve as vice-leaders who are interested in improving collaboration with pediatric department chairs. To improve gender diversity, the SPR might consider sponsoring a "50/50" Postgraduate course in which each lecture topic is split between a male and female presenter. Purpose or Case Report: Common cause analysis of exams actively submitted as learning opportunities in a peer collaborative improvement (PCI) process can gauge both opportunities to improve and potential risk to patients. Here, rates of learning opportunities were evaluated based on radiologist assignment, patient type, and exam priority.

All exams dictated at our pediatric hospital between July 3, 2016 (first full day 24/7 in-house attending) and July 31, 2018, were obtained from RIS including time of dictation, patient type, and exam priority. Exams were categorized based on radiologist assignment, patient type-Inpatient (IP), Outpatient (OP), or Emergency (EC), and exam priority (Stat, Urgent, and Routine). Assignments were defined by institutional rotations of Evenings (5p-10p), Nights (10p-7a), and Daytime (7a-5p, analyzed in two groups as Weekdays and combined Weekends & Holidays). Actively submitted PCI learning opportunities were identified and categorized. Rates were calculated by dividing number of learning opportunities by total number of exams in each category. Chi-square test was used to analyze rate differences. Pairwise comparisons were made for rotations and patient types with Bonferroni method adjusted p-values. Note that the peer learning system studied here focuses primarily on the professional aspects of radiology, and is separate from the hospital incident reporting system which was not evaluated here. 

In this study, the highest learning opportunity rates occurred on evenings and inpatient studies. These higher rates correspond to increased risk and are likely related to both high volume of studies per radiologist (evening assignment) and exam complexity (inpatients). Purpose or Case Report: Teaching files (TF) are often underutilized at institutions for various reasons including inefficient software programming, poorly designed user interface, or medical staff resistant to change. We demonstrate how to repurpose a teaching file in order to bolster user acceptance and educational value by optimizing TF integration.

Using a pre-existing teaching file in the Primordial® platform which was available but rarely used, we repurposed the program to improve functionality and user interface as well as improve ease of use and encourage integration. We demonstrate an all-encompassing 6 step approach to curriculum development based on Kern including identifying the problems, performing a targeted needs assessment, outlining goals and objectives, providing a vision for the TF, and listing desired features. Multiple 5-minute information sessions were implemented at the start of the new trainee year to reinforce use prior to educational conferences. A hands-on simulation session for all residents and fellows was given by an information technologist specializing in radiology and Primordial®. Finally, we reinforced educational strategies by requiring the TF to be incorporated into interdisciplinary conferences given by attendings and fellows. Results: Our all-encompassing approach was executed to improve functionality and increase use. Teaching File integration into the program was successful with a significant increase in utilization since its implementation. The number of teaching file entries were recorded at zero in March 2018, and after implementation of the new TF, teaching file entries steadily increased (14 -June 2018, 18 -July 2018, 34 -August 2018) with continued upward trajectory. Conclusions: Through a deliberate and thorough process, a fully integrated, repurposed Teaching File was created and successfully integrated into the radiologists' daily routine.

Radiologists at all training levels now incorporate TF into lectures, teaching file educational lectures, and interdisciplinary conferences and catalogue individual cases. Purpose or Case Report: RADIAL LMS, a tool to manage and increase timely access to educational content that tracks utilization and progress over time, was launched in October 2017 to consolidate valuable resources to enhance radiology education and the quality of work at the PACS. We present the utilization and perceived value of this tool one year later. Methods & Materials: RADIAL LMS was created following a methodical educational research process that included: needs assessment; inventory of existing resources; analysis and choice of best platform; implementation readiness assessment; and roadmap. A timeline for staged rollout and scalability beyond our institution was also developed. The LMS was launched Oct 2017 to all radiology trainees and attending radiologists holding institutional sign-on. Access was also given to administrators, researchers, assistants and incoming fellows. LMS analytics and focus group interview of current fellows and residents were tools used for assessment and evaluation of the LMS including: Navigation: number and time of access, number of users, top resources accessed; Rate of development of courses and curriculum; Appeal: look and feel of user interface, ease of navigation, preferred resources, drawbacks, and future directions for improvement and expansion. Results: Navigation: 178 registered users with 40-79 logins per day; 1-20 logins per day 2000h-0730h ; 1000 logins per month; average time per resource 2-4 minutes; top "courses": toolboxes and research onboarding. Course Development: 261 courses and 11 curricula; three apps; 51 new objects per month. User Interface: Appealing and intuitive interface but inconsistencies with access related to firewall issues at the institution level.

Most helpful for toolboxes, presentation preparation, accessing papers, on-the-go learning. Future directions: Ongoing program assessment with seamless integration with PACS. Process mining using deep learning to generate student modeling algorithms to drive adaptivity and produce true mastery. Interinstitutional collaboration to create a universal resource for pediatric radiology. Conclusions: RADIAL is a full-service LMS that has gained traction as a resource used for working, teaching and learning at point of care and beyond. The largest drawback related to inconsistencies in access has a manageable solution. It has future promise as a game-changer to support the breadth of subspecialty radiology knowledge to maintain mastery-level image interpretation and promote mastery of the field for trainees. Purpose or Case Report: To foster a global quality culture, many hospitals deployed informatics and organizational solutions, including daily readiness huddles to track department operational issues, reporting systems to track enterprise wide safety and near-miss events, and statistical process measures for quality improvement. Radiology visits have a unique daily workflow encompassing Planning, Acquisition, Post-processing, Interpretation, and Communication (PAPIC). Current radiology information systems (RIS) do not capture opportunities for improvement in the above spheres, which are patient encounterspecific. Therefore, error detection and reporting is unreliable. Capturing this detail is critical in identifying and tackling improvement projects that have direct impact on patient care.

In the radiology department of a large pediatric center, we created an informatics solution called OPEN (Ongoing Professional Evaluation) within RIS to allow quick 15-30 second capture of real-time encounter-specific opportunities for improvement. Input is categorized based upon imaging workflow (PAPIC), with a free-text comment box for radiology experts to describe issues. Clinical specific data is automatically captured in the hospital information system (HIS). The imaging quality team further augments data collection by categorizing into granular quality elements to evaluate for themes and improvement opportunities, which are then communicated by weekly reports to section leadership for resolution. The collected information is deemed confidential and privileged peer review data. Results: This new mechanism was envisioned in 2017, built and implemented in the summer of 2018, with staff training and education prior to launch. Engagement and trends are being tracked and evaluated from July-Oct 2018. A total of 385 entries have been reported in OPEN. Of the 385 recorded entries, the largest category was acquisition issues (51%). Acquisition issues include subcategories such as artifacts, motion, technique, wrong exam, equipment errors and missing information. The second largest category was planning. This commonly included issues obtaining correct orders or protocols which lead to delay in patient care. Conclusions: OPEN is created as a non-punitive, user friendly, quality reporting tool in radiology. It seamlessly integrates with RIS/HIS to provide automatic contextual information, and causes little disruption to patient care, thereby promoting engagement and insight into opportunities for improvement in imaging. Despite similar noise-levels (p=0.38, CNR was significantly higher (p<0.05) in low weight-groups (W1 15.1; W2 11.1) compared to high weight-groups (W3 9.8; W4 6.7; W5 6.0). Conclusions: Evaluation of the RCR-adapted pediatric trauma protocol showed relatively poor vascular and solid-organenhancement in the highest weight-group and high vascularenhancement in the lowest weight-group. Spleen-homogeneity and solid-organ-enhancement were insufficient in a substantial number of cases, mainly in higher weight groups. Our study indicates that tailoring unified protocol for different weight groups is challanging and further refining of the split-bolus single-pass CT protocol for children is essential. Purpose or Case Report: A number of upper gastrointestinal (UGI) studies were performed at our institution and interpreted as indeterminate for UGI rotation. These indeterminate cases led to family anxiety and provider uncertainty about the best course of action. We undertook a retrospective case review of all UGI studies performed over a 1 year period as part of a quality improvement initiative. The purpose was to identity the percentage of indeterminate UGI cases and any opportunities for improvement in imaging technique that could reduce this rate.

All upper UGI studies from June 2017 to May 2018 were identified using the electronic medical record. All cases were categorized as normal, abnormal, or indeterminate rotation. All indeterminate studies were reviewed by a board certified pediatric radiologist using a list of criteria for optimal UGI technique:1. Avoid over distension of the stomach with air or barium2. With true lateral position, document the retroperitoneal position of the 2 nd and 4th portions of duodenum3. Use an unrotated AP view to document the duodenal jejunal junction (DJJ) to the left of the left pedicle of the spine, and at the level of the duodenal bulb4. If there is uncertainty regarding position of the DJJ, follow the barium to cecum or consider contrast enema. Results: 514 UGI studies were performed over the 1 year period, with 24 dictated as indeterminate for UGI rotation. 6 were excluded because the duodenum did not fill with contrast (2 pyloric stenosis, 2 refusal to drink, 1 small bowel followthrough, 1 duodenal web). The remaining 18 cases (3.5%) were all indeterminate because of low position of the DJJ. With radiologist review for imaging technique, no technical issue was identified in 5/18 cases. Patient rotation in the AP projection (7/18) and/or lateral projection (4/18) was the most common technical factor. Overdistension of the stomach was seen in 4, and duodenal sweep was poorly seen in 2 cases. Contrast was followed to the colon in 4 cases, with normal position of the cecum in 1, abnormal position in 2, and poor visualization in 1.

Over a 1 year period, 3.5% of UGI studies were indeterminate for UGI rotation. Patient rotation was the most common factor that may have contributed to indeterminate impression. To help reduce this rate, future intervention will focus on the radiologist and technolgist working cooperatively to acheive optimal patient positioning and technique using a verbal checklist. Purpose or Case Report: The purpose of this study is to educate on the importance of proper patient positioning and immobilization techniques for the X-ray technologist while performing suspected non-accidental trauma cases. It will further explain the types of fractures that are typically seen in these cases and the subtlety of such fractures.

Non identified case studies, discussion of immobilization devices and discussion of vulnerable population will be utilized. Comparisons of confirmed cases of non-accidental trauma (abuse) to cases where abuse was ruled out. Explanation of techniques technologists can use to properly immobilize patients along with helpful positioning aids. Photographs of these devices will be provided. Results: Proper immobilization and positioning skills help to improve the diagnosis of non-accidental trauma cases. This allows those children that are most vulnerable to receive the medical care and social work intervention that they require. Conclusions: When patients are immobilized correctly and optimal positioning is utilized the best images possible can be reviewed by Radiologists. The Radiologists are given the best opportunity to detect very subtle fractures that are often present in non-accidental trauma cases. Proper diagnosis of nonaccidental trauma allows for child welfare agencies to utilize protocols to get these children proper care. Purpose or Case Report: The Children's Hospital of Philadelphia first started additive manufacturing i.e. 3D printing in 2012 and has since renamed its laboratory to the Children's Hospital Additive Manufacture for Pediatrics (CHAMP 3D). CHAMP 3D focuses on the use of 3D printing in surgical planning, educational training, and research. The following provides some operative insights into CHAMP 3D and some tips on starting an AM/3D lab.

Setup & LocationDetermining setup and location is a crucial starting point. An in-house lab allows for convenient collaboration within direct workflow of many groups, but this convenience also comes at a cost. The involvement of building operations is essential to ensure space is feasible.Goals & MachinePrinter selection is interlocked with the lab's goals and can set limitations on project scope. The first steps in establishing a 3D lab would be to outline the ideal set of projects and find a machine to match, taking into consideration examining material, cost, applications, and space availability.Team & CollaborationsIt is important to have a dedicated team of radiologists, engineers, administrators, and technologists. Each plays a key role in running the lab, brainstorming projects, segmenting scans, and expanding opportunities.Cost CenterThe CHAMP lab has introduced an internal billing model. Departments are charged for their prints based on a combination of material used, time processed, and flat overhead rate. The overhead rate has the dual purpose of recouping the cost of running the lab and limiting frivolous printing.OutreachOutreach can greatly extend the possibilities of the lab, ranging from collaboration to education and even recruitment of new team members. Collaborations with AM/3D labs can be an excellent opportunity for resource and knowledge sharing. Education and Maker events provide networking and discussion on the newest methods. Internship for students can expand the team and promote interest in additive manufacturing. Results: In the last year alone, the CHAMP 3D Lab has supported over twenty-five different groups across the hospital and university. A new metric of success targeting papers, intellectual property, and grants (PIG) has been introduced to drive the focus of the lab and project. Conclusions: Hospitals around the world continue to establish additive manufacturing labs for research and clinical applications at an ever increasing rate. Strong planning techniques can greatly improve the effectiveness and success of the lab. Purpose or Case Report: Phantoms and task trainers are utilized in a variety of training and education purposes. Handsontraining with realistic tools allowsfor a controlled learning setting and practice. Unfortunately, these phantoms and trainers can have prohibitive costs and tend to lack pediatric options. The growth of additive manufacturing (3D printing) and other fabrication methods can be used to create pediatric trainers to meet these educational needs unavailable on the market.

Additive manufacturing and other fabrication methods, such as silicon molding, are beneficial tools when designing and creating custom phantoms and trainers. Additive manufacturing models can be printed in various types of plastic, rubber, powder, or metal to create a phantom. It can also be used to print molds for silicon pours, therefore allowing for detailed forming. Silicon ranges in a variety of softness, durability, and other certified properties. The design for these models can be generated from a combination of scans that can be generalized and anatomized. By referencing the scans, models can be very realistic in detail. In addition, features can be added to facilitate the use of the models. Results: Through experiments with designing phantoms, a few tips have been developed to accomplish a variety of effects. First is the use of printed models for embedded structures. A printed bone structure can be encased in soft silicone to give a realistic feel and visualization on imaging. The embedded parts can also be painted to alter the exogeneity. Additives, such as sand or Metamucil, can be mixed into silicon to alter the texture and exogeneity, creating varied regions in scans.Internal cavities can be a challenge when molding but are essential in medical trainers. Break molding and dissolvable materials allow for these cavities to be as detailed as external surfaces. Both processes start by embedding printed cavities and pathways in poured silicon. In break molding, the internal parts are forcibly cracked then pulled and/or poured out. Purpose or Case Report: The purpose of this case is to analyze three main components for MR Appendicitis and MR Brain Ventricles. The first component is to evaluate the reduction of radiation dosage by changing the exams from CT to MRI. The second component is to evaluate the reduction of need for general anesthesia and/sedation. The third and final component is to evaluate the reduction of the overall cost.

The research is being conducted by utilizing two types of reporting systems EPIC and mPower. Our time frame for MR Appendicitis is from 3/10/2015-10/27/2018 and MR Brain Ventricles is from 04/27/2016-10/27/216. Within these time frames we are evaluating three main aspects; reduction of general anesthesia/sedation, reduction of cost and reduction of radiation. Our evaluation is taking the information within the timeframe that has been determined to evaluate the three components and calculate the end results. Results: 1. Reduction in overall costa. Cost analysis of MR vs CT examsb. Cost analysis of CT contrast and materials2. Reductions in General Anesthesia/Sedation medicationsa. Cost analysis of GA/Seditions vs savings3. Reduction in Radiation Dosagea. Analyzing the estimated reduction of radiation dosage altering the modality to MR Conclusions: Our evaluation on the improvement of Patient Care has enhanced by utilizing altering imaging methods. This is a substantial change in medical imaging that has been developing over time to have a positive impact on patient care. As we look at our statistical results it will conclusively show that the reduction in radiation dose has increase as well as the cost savings to the patient and the organization. It also will be able to prove that decreased need of general anesthesia/sedation that gives patients the ability to continue their health plan faster and less stressful.

Melissa Goehner, melissa.goehner@choa.org; Nikki Butler, BMSc, RT(R)(QM); Diagnostic Radiology, Children's Healthcare of Atlanta, Stone Mountain, GA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: To identify that engaging staff to be invested in the workplace and the importance of the work being done will increase the customer service initiatives and develop a more satisfied patient population.

There have been many changes implemented to increase staff engagement including changing from 8 hour shifts to 12 hour shifts across all modalities, bringing all modalities to full staff, new phone technology to include texting options throughout the hospital, including staff in decision making, parties and events( some competitive), leadership rounding and increased communication, staff projects and committees for staff to work together, regular leader meetings, Thank you cards, Dance cards to show patients who have multiple tests in multiple modalities which exams are first and to make sure that all tests are completed prior to leaving the facility, lunch and learns to educate staff and build relationships with the radiologists and administrative staff. Results: There was a boost in May of 2018 to focus on improving customer service. When improving staff engagement initiatives, customer service scores have been on a slow and steady increase since utilizing the methods and materials. There has been a 3% increase in the departmental customer service scores within the past five months. If staff engagement efforts continue, a steady increase in scores is expected. Conclusions: When staff are engaged in the work they do and invested in the facility, and believe that the organization is invested in staff, it results in customer satisfaction. Methods & Materials: IH-DCMRL involves ultrasound guided injection of a gadolinium contrast agent into the intrahepatic lymphatic ducts followed by MRI of the chest and abdomen with dynamic time resolved imaging. Initially a 25 gauge 3.5 inch spinal needle is placed intrahepatically adjacent to a branch of the portal vein under ultrasound guidance. Needle position is confirmed with fluoroscopy. The patient is then transferred to the MRI suite. Following a heavily weighted 3D T2 weighted sequence, gadolinium is slowly injected into the liver lymphatics and dynamic T1 weighted MR imaging is performed every 5-20 seconds over 6-7 minutes. This is followed by a high resolution respiratory navigated 3D IR inherent gradient echo sequence. Results: MRI assessment during gadolinium injection of the intrahepatic lymphatics, can be successful technique for evaluation of the liver lymphatics, thoracic duct, and to visualize bowel perfusion as seen in PLE and peritoneal perfusion as seen in chylous ascites. Conclusions: IH-DCMRL is a promising cross-sectional imaging technique to evaluate liver lymphatic flow and can be helpful for assessment of certain conditions involving abnormal liver lymphatic flow. Purpose or Case Report: Over the years, the ingestible electronic pill has been shown to be a clinically relevant technology platform, with biomedical applications such as drug delivery, temperature measurements, gastric emptying and transit times. At the same time there is work being done on the next generation of ingestible devices such as those powered by stomach acid or the Capsule Ultrasound device.

Methods & Materials: devices such as those powered by stomach acid or the Capsule Ultrasound device by Arbabian LabThe increase usage of the ingestible devices such as the PillCam, and Smartpill poses a challenge for us in MRI.Families currently have a hard time recognizing or identifying implanted devices but seem to be more challenge with identifying or recognize ingestible devices during our screening process.Most MRI screening process consist of a screening form and or interviewing process that has been designed to detect devices that are either wearables or embedded in the patientsWith the increase use and development of new ingestible devices, we must look at our current screening forms and interviewing process. Results: When scripting our interviewing process we try using questions that will help trigger for surgical and or wearable devices.As we all know the pressures of getting a diagnosis, as well as being a poor historian can sometimes interfere with how families respond to questions about devices but the connection between an ingestible device that is associated with the word pill becomes even more challenging.We need to continually change our MRI screening and interviewing processes with advancements in technology that can have an effect on being safe in the MRI environmentTo try and meet this challenge we had discussions with MR safety leaders, and decided on the following changes;Added the following change to our screening form, Ingestible(PillCam, SmartPill, etc.) with a Yes or No check boxScreening interviewing process,added these trigger questions;1. Do you or your child have anything in them that he or she was not born with?2. Have you or your child swallowed any devices for medical use?3. Have you or you chld been seen in the Gastro-Interstinal Clinic? Conclusions: Recent advancement in ingestible devices and advances for next generation devices has become a challenge to us in identifying ingestible devices. Changes in technology bring changes to the MRI safety screening process. Continual updates to both safety screening forms and the interviewing process should be reviewed as new devices are identified Purpose or Case Report: The goal of my talk is to present my audience with the obstacles that I faced as an MRSO when implimenting and executing new safety guidlines. I will also outline the safety team and explain their roles in our safety council. I will describe how we changed the over all culture in our department with the help of certain tools to accomplish our safety benchmarks. Methods & Materials: I will describe the process of implimentation of rules and guidlines that we crrenttly use in our MRI safety process. Show the steps which were taken to update and improve our MRI safety culture. Results: Explain how our safety process is better know thatn it was two years ago. Conclusions: MRI safety is a on going process that will constsantly change as a department grows in size and as its patient numbers increase. With that being said the rules and guidelines must change as well. Purpose or Case Report: One treatment for congenital scoliosis is the use of spine-straightening rods. Previously, the rods that have been surgically implanted were non magnetic and required the patient to undergo additional surgeries (usually every 6 months) to extend the rod length as well as radiographs pre and post surgical to evaluate the distraction. Currently many of the rods in use at our institution are MAGnetic Expansion Control (MAGEC®) rods that can be extended magnetically with the use of an external remote controller, thereby no longer requiring repeated surgeries every 6 months. As a result, patients have distractions performed every other month. While this is a marvelous improvement, these patients are now receiving an increased number of radiographs to confirm lengthening and interval correction of their condition due to the ability to distract every other month versus every 6 months with non magnetic rods. Several studies investigated whether ultrasound could be used as an alternative to X-ray when measuring rod lengthening in order to reduce radiation exposure. The result showed that ultrasound correlated very well with X-ray measurements and that while X-ray would still be needed, ultrasound could be utilized to eliminate the number of X-rays done and thereby significantly reduce radiation exposure to the patient. We were asked by the orthopedic department to see if we could institute a protocol for ultrasound imaging at our institution for these patients based on Cheung's study. What began as a QI initiative has evolved into an imaging clinic for these patients. We will show that the implemented standardized protocol and imaging clinic reduced unnecessary radiation, improved time management and increased the quality of care these patients receive. We intend to share the knowledge we have acquired of this successful learning model with other facilities around the country.

This study will look at the development of this clinic and evaluate its progression to include the length of exam times and the amount of radiation exposure between the standard x-ray method and the newer ultrasound method. Results: Our imaging clinic resulted in an 83% reduction in radiation and a 64% reduction in wait time to our patients. Purpose or Case Report: Ultrasound equipment and transducers can be intimidating to young patients. Many videos and graphics have been made for education, but can be perceived as disconnected from the clinical process. To help address these fears and engage collaboration between sonographers and their patients, a phantom was designed for children to play and learn before their own scan. This phantom would need to be pediatric-friendly, simple use, and educational.

The first step was to design an engaging, approachable figure that children would be interested in scanning. The CAD program SolidWorks was used to design a stout, one eye, spiked "monster." A mold frame was printed on a Fortus 450mc in ABS plastic. The embedded parts to be hidden inside the monster's belly included a crown, a heart, a cat, and a duck; Selected due to their simple shapes that children of all ages and backgrounds should recognize. These shapes were printed on a Connex 500 in VeroWhite plastic.The "monster" was molded out of silicon in a layering process. The silicon can be dyed any color as desired. The initial 0.5 inchlayer of silicon was poured. Once set, the shapes were placed perpendicularly in a pattern within the belly region. A 1.5-inch layer of silicon was poured on, covering the shapes. A fabric face was added and the final 0.5-inch layer was poured on top.

Once dry, the "monster" was pushed out of the frame and cleaned up. Results: Five sonographers were able to scan and find the various shapes using curved array C9-2 MHz and linear array 12-4 MHz transducers under pediatric abdominal and superficial presets. The technologists were challenged and successfully figured out most of the shapes. Notably the cat shape showed some challenge with approximately three misidentifications.

The "monster" phantom has since been used as part of the Ultrasound Awareness Day in our institution, as well as in our pediatric ultrasound department. In the clinical setting, this "monster" can be used prior to an exam and allows the pediatric patient to work with sonographers to understand the procedure and potentially help to calm their fears.

In summary, this pediatric-friendly "monster" phantom has been used in our institution to educate children and offers an excellent opportunity for engagement between the sonographer and the patient. The next step will be to work with the child life department to run a larger scale study evaluation of the "monster" phantoms and the potential benefits to pediatric patients. Purpose or Case Report: Ultrasound scanning of infant'ship can prove to be a challenge to learn for some sonographer from the precise angling of bones and handling of the not so understanding babies. Phantoms proved a calmer situation to learn and practice skills. To this end, an infant hip phantom with ultrasound visible fat, bones, and femoral head was created. The model will also need to simulate the physical aspect of the scan such as the motion and positioning of the hip and bones.

The fabrication of the hip started with the selection of sample patient scan. For this, it was determined to focus on 3-6-month-old patient with pelvis or adnominal CT scans with normal bone structures. The bones and body (skin and contains) were segmented out using Materialise Mimics. A mold was created with a subtraction of the body and pouring hole added. Posting was added to the bones to suspend within the mold and handle extension were placed to the end of the femur to represent the leg. The femoral head was designed geometrically and aligned it to the femur.The bones were printed on a Connex 500 in VeroWhite, which most resembled bone in hand and under ultrasound. The mold and post were printed on Fortus 450mc ABS plastic. To improve the echogenicity of the bones, it was determined to paint them with acrylic paints. The bones with femoral head and posting were assembled in the box and the skin toned dyed Smooth-on Ecoflex 10 silicon was poured.

The hip phantom was tested by several technologist and radiologist. The phantom was placed on the bed and was treated like a normal patient allowing sonographer to apply gel and use traditional set up to practice the procedure as if with a patient. Each user ran through the standard extent angles: transverse neutral, coronal neutral, transverse adduct, transverse abduct, coronal flex and coronal posterior lip.The phantom proved successful in imaging of the spine, pelvic bones and femoral head under scanning with a little noise seen in the silicon that was reduced with degassing process during pouring. Several commented on the realistic feel of manipulating the leg and the click of the joints. During one test a student technologist was even instructed in the probe placement and joint extensions. Conclusions: Moving forward the next steps will be to fully implement the phantom into the sonographers training and education. This phantom has lead to additional requests for other regions of the body, such as lumbar spine and neuroradiology applications. . In order to better and more accurately diagnose the severity of the venous compression, an intravascular ultrasound (IVUS) allows the operator a full 360 degree view of the vessel. An IVUS catheter has an ultrasound transducer attached to the distal end of the catheter. There are three sizes of IVUS catheters; each size allows a specific viewing diameter based on size of catheter. The catheter is advanced through the area of concern, and the images are recorded and reviewed on a separate workstation. Individual screenshots and diameter measurements can be obtained and saved. If a flow limiting stenosis is present, further treatment such as angioplasty or stent deployment can be pursued. IVUS is particularly helpful in these situations because it allows precise targeting at the area of stenosis as well as post intervention imaging to evaluate the success of angioplasty/stenting.

The use of an IVUS catheter is greatly beneficial in diagnosing, treatment planning, and post-intervention assessment of venous compression syndromes. 

The intervention proposed used a cloud based interactive audio/visual platform to facilitate communication between the radiology faculty and NICU treatment teams at their respective off site locations. The software platform was required to provide reliable, userfriendly, and seamless interactivity in order to maintain the same level of service currently provided by onsite consultation. The intended purpose of the intervention was to reduce travel time for pediatric radiologists. Analysis of the travel times was performed and time saved was determined based on preintervention travel durations. The preservation of consultation quality following intervention was a primary goal. To that end, the NICU clinical team members as well as pediatric radiologists were surveyed on subjective and objective measures of quality. Results: Travel times were significantly reduced following implementation of remote consultation at remote NICU facility, estimated to save a potential 4940 radiologist-hours over a six month timeframe. 99% of NICU providers responded very highly to questioning regarding the added value of a radiologist's presence at NICU rounds versus NICU rounds conducted prior to radiology's presence during daily rounds. Satisfaction results were positive after implementing radiologist's presence on NICU rounds. Conclusions: Feedback from pediatric radiology faculty and NICU treatment teams was widely positive following the implementation of telerounds. We demonstrated the ability to provide a similar level of quality of communication, ability to entertain dialogue regarding exams, timeliness of rounds from both surveyed groups. Survey of pediatric radiologists and NICU providers both show significant increase in satisfaction after implementation across all metrics. Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Gadoxetate disodium, a gadoliniumbased contrast material utilized in hepatobiliary magnetic resonance imaging (MRI), is associated with transient respiratory motion artifact during the arterial phase in adults. The purpose of this study was to determine if the transient respiratory motion phenomenon is observed in children.

This retrospective cohort study was approved by the institutional review board; informed consent was waived. Patients aged 4-18 years who underwent dynamic liver MRI with gadoxetate disodium between October 2010 and January 2018 were identified. 130 exams from 130 patients were selected for review, including patients imaged awake or under general anesthesia. Demographic, medical history, and imaging data were recorded for each patient. Three blinded reviewers scored respiratory motion artifacts on precontrast, arterial, portal venous, and late dynamic phase images using a 5-point Likert scale. Analysis of variance (ANOVA) was used to assess differences in average motion between phases in both the awake and general anesthesia cohorts; significant results were further evaluated using Tukey's multiple comparisons test.

Multivariable linear regression was used to identify significant predictors of arterial phase motion in awake patients. Results: 130 patients (65 boys and 65 girls; mean age: 9.8±3.7 years; awake, n=63; general anesthesia, n=67; gadoxetate disodium dose, 0.05 mmol/kg) were included. The anesthetized cohort was slightly younger than the awake cohort (8.9±4.0 vs. 10.7±3.1 years; p=0.007). There were significant differences between phases in average motion scores in the awake cohort (p<0.0001) but not in the general anesthesia cohort (p=0.051).

In the awake cohort the arterial phase motion score (mean score: 3.52±0.83) was significantly higher than the precontrast (mean score: 3.14±0.81; p=0.0003), portal venous (mean score: 3.07±0.92; p<0.0001), and late (mean score: 3.05±0.89; p<0.0001) phase motion scores. Age, sex, body mass index, presence of ascites, presence of pleural effusion, and total contrast dose did not significantly predict arterial phase motion score in the awake cohort. Conclusions: We observed significantly increased arterial phase respiratory motion artifacts in awake children undergoing dynamic liver MRI using gadoxetate disodium, suggesting that transient respiratory motion artifact does occur in children. A similar finding was not observed in patients imaged under general anesthesia, which may suppress this phenomenon. Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Liver transplantation is a potentially lifesaving treatment for pediatric liver failure. While some elements of perioperative liver transplant Doppler examinations are routine in most practices, there is no standard protocol employed across institutions. To date, no large-scale investigation has been performed to determine best practices based on a consensus of radiologists who perform pediatric liver transplant intraoperative Doppler (IOD) examinations. Our goal is to survey radiologists who perform these studies to better understand variability amongst institutions in the use of IOD in pediatric liver transplant patients.

With IRB approval, an online survey was distributed to 1600 members of the Society for Pediatric Radiology (SPR) via the email list serve. To avoid redundancy in responses from the same institution, participants were encouraged to identify one single person from their respective institution to respond to the survey. The name of the participants' institution was also requested to identify respondents from the same institution. Participants were excluded if pediatric liver transplantation was not performed at their institution, as well as if they were trainees. Results: Of the 1600 survey recipients, 50 (3%) responded to and completed it, representing 29 of 58 (50%) pediatric liver transplant centers within the United States. IOD was performed at 69% of pediatric transplant centers. In centers performing IOD, 47% performed examinations routinely with every pediatric liver transplant. The hepatic artery was the vessel most commonly routinely interrogated (87%), followed by the portal vein (83%) and hepatic veins (67%), although 50% reported that the exam protocol depended on the area of clinical concern. The surgeon and radiologist operated the transducer with relatively equal frequency. The pediatric radiologist rendered interpretation in 93% of institutions, although frequently in conjunction with the surgeon (30%). Surgeon preference (38%) was identified as the most common barrier to further expansion of IOD programs, followed by radiologist availability (25%). 47% of respondents deemed IOD "extremely useful" in the setting of pediatric liver transplantation. Conclusions: Despite the perceived utility of IOD in pediatric liver transplantation, there is considerable variability in the use of this technique. Our results support the notion of a team approach between surgeons and radiologists in further expanding the use of IOD in this population. Purpose or Case Report: We conducted this study to determine if there is an imaging biomarker associated with ALK mutations in patients with neuroblastoma(NB). MYCN oncogene amplification is identified in 20-25% of all NB cases and portends a poor prognosis. Aberrations of the ALK gene are also characteristic of the disease. Both are located on the p arm of the second chromosome, and are usually associated with aggressive disease and poor prognosis.

Data was collected from subjects under 12 years of age diagnosed between July 2008 to February 2018. Clinical, surgical, pathologic and genomic data were extracted from the medical record. Imaging reviewed by a single radiologist included MRI and I123 MIBG scans performed prior to surgical resection. Recorded imaging findings were site of primary lesion (adrenal, neck/thoracic apex, thoracic, lumbar paraspinal, pelvic, midline abdominal retroperitoneal and metastatic disease to lymph nodes, liver, lung, brain, and marrow). MIBG avidity and Curie score were recorded. Inclusion requirements were pathology results from primary specimen at diagnosis or after induction chemotherapy. Results: 66 subjects with mean age of 2 years, (59% female and 41% male) found 20% (N=13) were ALK positive. Of these, 6 were deceased with 4/6 also MYCN+. Additional mutations were present in all 13. 6 were INRGSS stage L2, 5 were stage M and one was L1. Imaging patterns described as (#ALK positive/total, %) were as follows: Neck/thoracic apex (4/4, 100%), bilateral adrenal (1/3,33%), multicompartmental (2/17, 28.5%), chest/thorax (2/9,22%), midline abdominal retroperitoneal (5/19, 26%), liver metastases (2/9, 22%). Curie scores ranged from 2-27. Of the 6 deceased, 5/6 had additional mutations, including P53 and TERT mutations. Other mutations labeled as having unknown significance were also present. Of all 66 subjects, 86% had other chromosome aberrations.

The strongest association of ALK mutation with a specific primary site was the combined neck/thoracic apex. The lack of a specific imaging biomarker may be related to genomic heterogeneity. Detection of ALK mutations at diagnosis holds promise for personalized therapy with upfront treatment with ALK inhibitors combined with chemotherapy, a current open clinical trial for high risk patients.

Authors are listed in the order provided. An author listed in bold identifies the presenting author. Purpose or Case Report: Lemierre's syndrome is an extremely rare condition characterized by initial oropharyngeal infection with development of septic thrombophlebitis and subsequently disseminated septic microemboli. The syndrome remains a disease of considerable morbidity and mortality. The incidence is approximately 3.6 cases per 1 million per year. It is commonly caused by gram-negative Fusobacterium necrophorum. However, less than a third of cases is brought on by other anaerobic bacteria. Here we present a rare case of a Lemierre's syndrome in a child caused by methicillinresistant Staphylococcus aureus (MRSA). A 4-year old male presented to emergency department with unremitting fever, progressive painful submandibular swelling concerning for Ludwig's angina, vesicular lesions on the skin, and decreased level of consciousness. The patient was reported to have fallen onto his chin with resultant lip laceration and tooth avulsion. Fever and neck swelling developed two days later. Computed tomography (CT) of the neck revealed findings in keeping with clinically suspected Ludwig's angina including: soft tissue gas, and diffuse fat stranding involving the sublingual, perioral, and right submandibular spaces with extension to right sternoclavicular muscle, carotid and jugular vessels. No abscess was identified but a focal non-occlusive thrombus was seen in the right internal jugular vein. Additionally, the lung apices demonstrated multiple patchy densities raising concern for Lemierre's syndrome and prompting further evaluation with a contrast-enhanced chest CT. Chest CT confirmed the diagnosis demonstrating multiple, variable-sized, randomly distributed lung nodules with cavitation, and multifocal consolidation consistent with septic emboli. Blood culture and skin swab of vesicular lesions were positive for MRSA. On further discussion, it was revealed that the patient's father was recently treated for MRSA abscess. The patient was placed on intravenous antibiotics (vancomycin, rifampin, meropenem) and anticoagulants (tinazaparin) with improvements in symptoms and imaging findings within six weeks post admission.The learning points include:1) the need to critically evaluate lung apices and vasculature on neck CT in patients with evidence of soft tissue neck/oropharyngeal infection2) despite the rarity of Lemierre's syndrome, multiple cases caused by MRSA have been described. Purpose or Case Report: Infantile myofibromatosis is a rare condition consisting of benign fibrous tumors typically deposited in the skin, soft tissues, muscles, bones, and visceral organs. The entity can be solitary or multicentric. Although controversial, outcomes are generally worse in cases with visceral organ involvement. The prognosis is generally favorable in cases that lack visceral organ involvement, with a majority of cases showing spontaneous regression. The imaging findings of infantile myofibromatosis will be illustrated using both prenatal and postnatal imaging including ultrasound, MRI, radiography, CT, and bone scintigraphy. We also present gross specimen and pathology images. Our case involves a 33 week 4 day gestational age male fetus that initially revealed dilated loops of bowel on ultrasound. Fetal MRI was performed at 34 weeks and 4 days, which demonstrated the dilated loop was colon in the region of the hepatic flexure. Additionally, multiple solid-appearing lung masses were noted, which had not been visualized on ultrasound. Because of concern for a possible metastatic process, the entire fetus was thoroughly imaged, but no primary source was found. The differential diagnosis that was given on the fetal MRI included metastatic disease from the mother or fetus, infantile myofibromatosis, or infectious etiology. The mother underwent induction of labor at 35 weeks 4 days, and the baby was born via uncomplicated vaginal delivery. To exclude transplacental metastases, the mother underwent dermatologic skin check, mammography, colonoscopy, and head CT, all of which were negative. Postnatal radiographs of the baby revealed a focally dilated loop of bowel, and the patient underwent exploratory laparotomy. In the OR, nodules were noted on the small bowel serosa resulting in a bowel obstruction. Chest radiograph and CT also confirmed the presence of multiple solid lung masses. Bone scintigraphy was negative. Pathology from one of the bowel nodules revealed infantile myofibroma. Our patient has not undergone any therapeutic treatment, and follow up imaging has demonstrated continued spontaneous regression of the lung masses. Although infantile myofibromatosis is a rare entity, it is important to include in the differential diagnosis in a fetus with multiple solid-appearing lung masses. Throughout the clinical course of these patients, imaging plays an imperative role in the assessment of these lesions.

Peeing Purpose or Case Report: Caudal duplication syndrome is a rare entity that involves duplication of various structures arising from the embryonic cloaca and notochord. This presentation outlines the case of a 38-year-old G2P1 woman referred to our institution at 21 weeks gestation for anomalies detected on antenatal ultrasound. Antenatal ultrasound demonstrated sagittal duplication of the bladder, duplex phallus, a bifid scrotum and a horseshoe kidney. A subsequent fetal MRI confirmed those findings and also demonstrated apparent duplication of the colon and dysmorphic lumbosacral spine. Following delivery, the baby passed urine via both urethras and meconium via a right-sided anus; there was a left-sided anal dimple with an imperforate anus. VCUG showed no communication between the two bladders. A colovesical fistula was demonstrated between the left bladder and colon, which is likely the redundant duplicated colon with the imperforate anus. MRI of the abdomen and pelvis showed a lipomeningocele with attempted sacral duplication, redemonstrated duplication of the pelvic organs, and confirmed absence of a left-sided rectum and sphincteric complex. Caudal duplication syndrome is a complex malformation, the management of which often requires a multidisciplinary approach involving radiology, general surgery, urology, and neurosurgery. This case illustrates the findings of this rare entity with good correlation between fetal and postnatal imaging. It also highlights respective contributions of the various imaging modalities in guiding management, which usually entails staged surgical correction. CT showed the "hyperattenuating rim" sign. This report is the first reported female pediatric case of FFLAT that was diagnosed with US and CT, given a trial of analgesics, and definitively cured via minimally invasive surgical excision. A 13-year-old female presented to the emergency department with episodic waxing and waning abdominal pain for three days. The pain had localized to the mid-epigastrium and worsened with deep inspiration. On examination, vital signs were within normal limits. Focal tenderness was elicited upon palpation of the epigastrium. Laboratory evaluation revealed a mild leukocytosis 14.2 x 109/L, normal range (3.9-10.6 x 109/L). Results: Ultrasound demonstrated an irregularly marginated hyperechoic mass in the epigastric region that extended into falciform ligament. Confirmatory CT scan of the abdomen/pelvis with IV and oral contrast revealed mass-like stranding of the intraperitoneal fat in the epigastric region with a classic ovoid hyperattenuating rim sign. The vessels extending into the falciform ligament did not demonstrate contrast enhancement further raising concern for torsion. Surgical consultation was obtained. Conservative management was recommended via a trial of non-steroidal anti-inflammatory medication. Despite several weeks of treatment, the patient's pain persisted without significant improvement. Transumbilical laparoscopic excisional findings strongly suggested an acute on chronic inflammatory process confined to the fatty appendage of the falciform ligament. The pathologist described the resected mass as fibrofatty rubbery tissue. The pre-operative diagnosis of FFLAT was confirmed. The patient was found to have complete resolution of pain at her two-week post-operative follow-up.

Conclusions: This case reiterates the clinical multidisciplinary team approach required to diagnose the rare case of FFLAT in a pediatric patient. It also demonstrates careful observation of conservative management, and optional minimally invasive surgical resection for patients with persistent symptoms.

Jane Tong 2 , Yu Luo 1 , yu.luo@vanderbilt.edu; 1 Vanderbilt University, Nashville, TN, 2 Drexel University, Philadelphia, PA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Although acute appendicitis is thought to be result from luminal obstruction of the appendix, rarely it may develop following abdominal trauma. Traumatic appendicitis is thought to occur through direct injury to the appendix, or as a response to other abdominal organ injury. On the other hand, in patients with other organ injury, some distension of the appendix can occur with surrounding free fluid secondary to trauma, mimicking appendicitis. While the clinical presentation of traumatic appendicitis is similar to that of traditional appendicitis, differentiation between reactive appendiceal changes in the setting of traumatic injury to other intra-abdominal organs is important, as the latter will not require appendectomy. We present two pediatric patients in whom following initial suspicion of acute appendicitis, ultrasonography (US) identified mildly enlarged fluid-filled and hyperemic appendix with out of proportion complex fluid, raising the suspicion of previously unsuspected abdominal trauma. Upon further examination, injury to other abdominal solid organs was discovered as the primary cause of patient's presentation and appendiceal findings were reactive to abdominal solid organ injury. In cases of suspected appendicitis, visualization of significant free fluid with dense debris on ultrasonography (US) calls for more careful examination to assess clues of other abdominal injury. Purpose or Case Report: We report a case of a 7 year old female who presented to urology clinic due to recurrent urinary tract infections that had started 4 years ago. The patient also reported symptoms of urge incontinence and nocturnal enuresis beginning at the same time. She previously consulted an adult gynecologist, which showed no physical exam evidence of genitourinary abnormalities. An MRI of the abdomen and pelvis was also ordered at that time which reported a normal exam except for a small left renal cyst. At our institution she underwent DMSA renal scan, which was normal. She underwent a voiding cystourethrogram (VCUG), which showed no vesicoureteral reflux. However, during the VCUG, an incidental note was made of large amounts of vaginal reflux extending into the cervix, uterus, and with spillage into the peritoneal cavity presumably via the salpinges. Conclusions: Extension of contrast into the peritoneal cavity via vaginal reflux on VCUG has previously been reported only in patients with genitourinary anomalies. However, this patient had prior workup demonstrating no definite genitourinary abnormalities. This case suggests that reflux of contrast to the peritoneal cavity may be possible in patients with otherwise normal genitourinary anatomy. Purpose or Case Report: Extramedullary hematopoiesis (EH) is defined as hematopoiesis occurring in organs outside of the bone marrow. It occurs in diverse conditions, including fetal development, normal immune responses, and pathological circumstances. These sites of extramedullary hematopoiesis may present as masses mimicking malignancy or produce symptoms due to pressure effects. In the setting of an existing malignancy they may appear as metastatic deposits signifying progression of disease. It is essential to confirm this due to its prognostic and treatment implications.We report a 2-year-old little girl who presented initially with an acute history of ataxia, nystagmus, tremor, mydriasis and bruises on her left forehead. A solid left suprarenal mass was detected and a diagnosis of Stage 4 Neuroblastoma and Opsoclonus-Myoclonus syndrome was established. Subsequently she was on treatment which included chemotherapy, IVIG and stem cell transplant. On an MRI of the abdomen done a year later, a single lesion was detected in the right lobe of the liver. On subsequent short term follow up, innumerable scattered lesions were seen in the hepatic parenchyma and were thought to represent metastases.

A liver biopsy showed that these hepatic lesions represented sites of extramedullary hematopoiesis.Extramedullary hematopoiesis has been uncommonly seen in the cranium and sacrum in the setting of Neuroblastoma. We believe this is a unique presentation with extramedullary hematopoiesis presenting as solid liver masses masquerading as metastases in a known case of Neuroblastoma.

Christopher J. Purpose or Case Report: Introduction: The management of gastrobronchial fistula (GBF) is not well defined in the literature. First line management is non-operative, allowing time for the fistula to close naturally. Surgical intervention is implemented when non-operative management fails. In medically complex patients, who often fail non-operative management, surgical procedures for closure remain high risk for complications.Case: We present a 2 year-old female with congenital esophageal atresia, duodenal atresia and annular pancreas. Beginning October 2015, she underwent multiple surgical procedures resulting in a multitude of complications. In August 2017, she transferred to our institution for management. Additional immediate operative intervention was pursued given the presence of the GBF. Despite two attempts at operative repair, the fistula recurred. Interventional radiology, consulted March 2018, developed a plan to place a percutaneous pigtail catheter in the esophagus through the dehiscent stomach wall from an external chest tube entry site, creating an iatrogenic enterocutaneous fistula (ECF). Diversion of flow from the GBF to the ECF was hypothesized to allow a conservative alternative for closure and tissue healing to occur. Initially, a 12 French (F) pigtail catheter was placed through the gastric wall into the esophageal pouch allowing the formation of the ECF tract. The patient returned for catheter downsizing to 8F three weeks later. A third intervention was performed to reposition the catheter to ensure that the tissues remained as dehydrated as possible. In May 2018, a 7 mm Cook Biodesign SIS fistula plug was placed. The 8F catheter was used to place a guide wire through the tract, then removed. The AFP bioplug, soaked in contrast to facilitate use of fluoroscopy, was placed through the mouth to seal the gastric wall. Placement was confirmed with a rigid esophagoscope and fluoroscopy. One week later, esophogram confirmed no leak. At 1 month follow-up the patient had no evidence of right pleural fluid and remains without evidence of fistula at five months. The skin site is well-healed.Discussion: In this case, an ECF was closed with a Cook Biodesign SIS fistula plug allowing for a successful non-operative strategy after multiple failed operative revisions. The plug has FDA approval for the treatment of anal fistula, but given the positive outcome in this case, its use in esophagocutaneous fistulas may be a viable off-label option for other patients. Purpose or Case Report: Leptomeningeal melanocytosis is a rare proliferation of melanocytes in the arachnoid and pia mater that presents as diffuse leptomeningeal enhancement. Findings in a rare case of this disorder are discussed along with differential considerations and diagnostic implications.

A 9-year-old male presented to the emergency department after hitting his head on the door of a car. CT of the head performed at that time showed minimal high attenuation material throughout the right parietal cortical reported as a small amount of subarachnoid hemorrhage. After it became clear that the patient was having uncontrolled seizures that may have contributed to the original trauma, he was placed on anti-epileptic medication and referred for MRI. Results: Multiple MRI's were performed with and without contrast over the next 6 months. The findings on the CT were found to be due to thickening of the leptomeninges which exhibited robust contrast enhancement, greatest in the right parietal lobe. Multifocal hemorrhages were observed in the left cerebral hemisphere over the course of the exams. A diagnosis of Sturge-Weber Syndrome was assigned. Since the leptomeningeal enhancement showed progression over 5 months, a surgical biopsy was performed. Upon entry into the calvarium, the surface of the brain was found to be dark brown in color. Biopsy revealed the diagnosis of leptomeningeal melanocytosis. Conclusions: Leptomeningeal melanocytosis is a rare disorder that can be confused with infectious, inflammatory, and/or vascular abnormalities. The disease is typically fatal with a short life expectancy following diagnosis. Careful attention on follow-up of leptomeningeal enhancement is essential to exclude a proliferative or neoplastic process. Purpose or Case Report: Ectopic thymic tissue may be found in the neck in up to 20% of the general population. Intrathyroidal thymic rest has been described as a rare entity, present in as many as 1% of children. The course of thymic migration parallels the thyroid and parathyroid glands, which explains their similar ectopic locations. While the natural history of these lesions has not been well studied, it is likely that there is involution of thymic rests with age. A few individual case reports have described the imaging features of intrathryoidal thymic rests confirmed by histopathology and flow cytometry. The aim of this case series is to raise awareness of the characteristic sonographic appearance of this entity to the radiology community.

This case series will review clinical and imaging features of 4 biopsy-proven intrathyroidal thymic rests. Characteristic imaging features will be elucidated. Differences between this entity and more concerning thyroid nodules that require biopsy and perhaps excision will be highlighted.

The presence of dot-dash echogenic foci within a focal hypoechoic nodule are characteristic for thymic rests but can be mistaken for more worrisome calcifications in a thyroid nodule. Additionally, the border can be lobulated and somewhat irregular, which is another significant feature of suspicious thyroid nodules by TiRADS criteria.

The sonographic appearance of thymic rests within the thyroid gland is very characteristic and should be suggested in the differential diagnosis of an incidentally found characteristic nodule in an otherwise healthy child. These nodules do not follow TiRADS criteria for benignity, however, radiologists should be aware of this entity. When patients have this classic appearance , conservative management should be considered, with less invasive sampling with fine needle aspiration favored over surgical excision if there remains clinical concern.

Poster #: CR-014

In Abnormalities on Chromosome 2p23 with a rearrangement of the ALK (anaplastic lymphoma kinase) locus causes abnormal tyrosine kinase receptor expression. Chromosomal abnormalities suggest a clonal origin and not just a reactive process or "pseudotumor," as these masses have been categorized in the past. Up to sixty percent of inflammatory myofibroblastic tumors express ALK which may help establish the diagnosis of the inflammatory mass as an IMT. The masses can occur in a variety of locations and have non-specific imaging findings, which will be reviewed.We present three cases, each in a different location:Case 1: 7 year old female presents with chronic cough and persistent right middle lobe abnormality on chest x-ray with concern for pneumonia or inhaled foreign body. She was found to have a soft tissue mass in her right mainstem bronchus with extra-luminal extension. CT, MR and gross surgical photos of the endobronchial IMT will be presented.Case 2: 20 month old male presents with hepatomegaly, jaundice and elevated bilirubin, alkaline phosphatase and liver function tests. Ultrasound and MR images of the pancreatic head IMT causing biliary obstruction will be presented.Case 3: 13 year old male presents with back, leg and pelvic pain. MR, CT and gross surgical photos of the right posterior pelvic sidewall IMT, which had evidence of nerve entrapment at biopsy, will be presented.The imaging, pathological and surgical findings from these patients, where available, will be presented and reviewed.We suggest that the radiologist consider Inflammatory Myofibroblastic Tumor in the differential diagnosis for inflammatory and fibrous lesions in children. There are several inherited disorders and congenital defects that predispose to progressive dilation of the thoracic aorta in childhood. Well known disorders like Marfan syndrome and bicuspid aortic valve have long served as clinical models but many, seemingly disparate, disorders result in similar abnormalities of the aorta. As knowledge about the genetic and histopathologic underpinnings of TAA grows, management will be guided by multidisciplinary teams. In order to maintain value in these teams, the pediatric radiologist must understand the varied causes of TAA, the underlying histopathology, and the treatment implications. This exhibit will introduce these predisposing conditions, review the similarities and differences, and review treatment options.

The following conditions that predispose to aortic aneurysm will be reviewed:I. Diagnosis has historically been made with fetal echocardiography, and palliative surgery performed frequently in the neonatal period. Cross-sectional imaging is typically reserved for post-operative complications. However, due to lower dose radiation and faster scanners, preoperative CT angiography imaging is becoming more common with the increasing use of cardiac EKG-gated CT angiograms, requiring pediatric imagers to be familiar with the diagnosis of the truncus arteriosus spectrum.

This retrospective imaging review will describe the imaging features of truncus arteriosus, including the 2-D multiplanar reformatted images and 3-D volume rendered images. Various anatomical types of Truncus Arteriosus detailed in the widely used Van Praagh classification will be demonstrated. Illustrative examples will also detail variants not clearly defined by the current classification systems. Important imaging distinctions between truncus arteriosus and other similar appearing congenital cardiac anomalies, such as Tetralogy of Fallot or pulmonary atresia with major aorto-pulmonary collateral arteries, will be described.

Conclusions: This educational exhibit will provide viewers with a framework to recognize and classify truncus arteriosus, as well as provide vital distinction between truncus arteriosus and other similar appearing conotruncal anomalies. Purpose or Case Report: Congenital heart diseases often occurs secondary to a variety of insults and rotational errors during development in utero. These can range from a spectrum of simple to more complex pathologies including arterial switches. It is important for radiologists to recognize the embryology and complications of arterial switches to further management. Today largely secondary to new advances in technology both corrected and uncorrected arterial switches are diagnosed more commonly; previously many patients with arterial switches may not have survived into adulthood.In utero, the primitive truncus is normally positioned anterior and midline. It eventually divides into the aorta and the pulmonary artery, which then rotates clockwise 150 degrees such that the pulmonary artery lies anterior to and left of the aorta. When variations occur it can result in a congenitally corrected transposition such as L-transposition of the great arteries (L-TGA) where the two ventricles are morphologically switched in position. With L-TGA, the truncus rotates 30-degrees clockwise which results in the aorta being anterior and leftward in relation to the pulmonary artery.In uncorrected transposition (R-TGA) the aorta arises from the right ventricle and the pulmonary artery arises from the left ventricle, secondary to a 30-degree counterclockwise rotation of the primitive truncus, the aorta is then located rightward and anterior to the pulmonary artery. This condition needs to be corrected surgically due to cyanosis resulting from the right ventricle not being able to supply the systemic circulation. Other variations of arterial switches that can occur is situs inversus where the rotation of the aorta and pulmonary artery is completely opposite of that which would be considered normal. Lastly in truncus arteriosus, the primitive truncus does not divide into a separate aorta and pulmonary artery. We will provide several cases from our institution of arterial switches on imaging as well as diagramatic representative models describing the embryology of how arterial switches and rotations occur. We also discuss the repairs and complications of these cases. Conclusions: Congenital arterial switches and complications from repairs can be confusing but are becoming more commonplace. It is thus important for radiologists to understand the embryology and identify the imaging characteristics.

Purpose or Case Report: Capillary malformationarteriovenous malformation (CM-AVM) is an autosomal dominant disorder with variable phenotype caused by heterozygous inactivating mutations in the RASA1 gene located on chromosome 5. Clinical manifestations are variable with cutaneous multifocal capillary malformation associated with fast-flow lesions. Most of them are located in soft tissues (intramuscular, intraosseous, spinal or cerebral). Many authors reported the clinical spectrum or the genetic association but few data are available on the imaging characteristics or criteria to establish the diagnosis of AVM. The goal of this poster is to review the imaging characteristics in the RASA1 series of our institution and in particular to evaluate the distribution of patients having true AVMs versus capillary hypervascularity.

A retrospective study of clinical and imaging files was conducted for all patients seen in our vascular anomalies group with a genetic diagnosis confirmation of RASA1 mutation. Institutional approval was obtained from the IRB and all patients (or parents) signed an informed consent. We reviewed 9 cases (8F, 1M) with RASA1 mutations. Cutaneous capillary malformation aspect, color Doppler ultrasound and MR findings will be described. Lung AVM has never been reported in RASA1 except in our series. Conclusions: Patients with RASA1 mutations have a wide spectrum of clinical manifestations. CM is the clue for the diagnosis, particularly if the CM has the pale halo. However, the AVM criterion for extracranial lesions has to be clarified. Most lesions in our series behave more like a capillary hypervascularization without AV shunting. Thus, the presence of extracranial/spinal true AVMs seems to be rare. The integration of genetic, clinical and imaging findings is important to have a better understanding of the disease and to offer the best treatment.

Jamie Frost, DO,j_mel46@yahoo.com; Michigan State University, Grand Rapids, MI Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Congenital heart disease (CHD) is the most common type of birth defect; affecting ~1% of the births per year in the U.S. Advancements in medical and surgical treatment have markedly improved survival and even infants with complex CHD survive into adulthood. However, extracardiac complications in the newborn period can cause increased morbidity and mortality. These complications can relate to alteration in flow dynamics, treatment changes, and/or sequelae of associated syndromes (Trisomy 21, 22q deletion syndrome, Heterotaxy, PHACES). The purpose of this exhibit is to highlight extracardiac complications of CHD and to review their imaging findings. Imaging findings to be reviewed will include catheter thromboses, necrotizing enterocolitis, arterial ischemic strokes, infections, malrotation, pulmonary complications related to prematurity, prolonged intubations, and associated syndromes, and complications of extracorporeal membrane oxygenation. The goal of this exhibit is to familiarize radiologists with the multiplicity of extracardiac complications in infants with CHD and to review the common imaging findings.

The Purpose or Case Report: Fetal MRI is now an important adjunct imaging modality in the evaluation of complex fetal anomalies, including cystic renal disease. The improved resolution and anatomic detail of the renal parenchyma offered by MRI can assist with identification, localization, and characterization of cystic lesions which are less clearly visualized on ultrasound. Advanced imaging adds value through enabling prenatal prognostication and patient counseling. In this presentation, we evaluate the spectrum of renal cystic abnormalities at the microscopic and macroscopic scale, and review the patterns of disease by cyst location, effect on parenchymal integrity, and obstruction of the collecting system. A clearer understanding of the diverse appearances and broad spectrum of outcomes of these fetal anomalies can contribute to more detailed treatment planning and more precise, familycentered care.In this presentation, we review the varied patterns of cystic renal disease as identified on fetal MRI, highlighting those forms known to be more associated with perinatal morbidity and mortality. Further, we correlate their appearance on MRI with pre-and post-natal US imaging, as well as pathologic findings. Finally, we will describe secondary prenatal imaging biomarkers that may be valuable in counseling and also with both definitive and palliative surgical planning.

Contrast-enhanced ultrasound for the evaluation of the neonatal brain: Diagnostic methods and scanning protocol Kayla Cort, DO, kayla.cort12@gmail.com; Maciej Piskunowicz, MD, Misun Hwang, MD; Children's Hospital of Philadelphia, Philadelphia, PA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Contrast-enhanced ultrasound (CEUS) for the evaluation of the neonatal brain provides additional diagnostic information when compared to conventional gray scale ultrasound through the detection of perfusion abnormalities associated with injury. When compared to cross-sectional imaging, CEUS has many advantages given its relative low cost and ability to be performed at the bedside, without the need for sedation or exposure to ionizing radiation. Diagnostic information is yielded through the qualitative evaluation of parenchymal enhancement patterns as well as quantification of microbubble perfusion kinetics from which time intensity curves are derived and additional perfusion parameters can be extrapolated. Given the 2-dimensional nature of the modality, a strategically designed scanning protocol is necessary to obtain the aforementioned quantitative values. The purpose of this exhibit is to demonstrate the current understanding of brain CEUS and educate on the brain CEUS protocol used for the diagnosis of neonatal brain pathology. Purpose or Case Report: Congenital eye malformations are relatively rare. However, they are often associated with complex clinical syndromes that require extensive prenatal evaluation and counseling. The purpose of this study is to provide a review of various congenital ocular anomalies that can be detected on prenatal ultrasound and/or MRI in isolation or as part of a syndromic findings and to address the importance of the prenatal genetic evaluation and parental counseling.

A retrospective review of the radiology database at our institution is performed. Various congenital orbital malformations are reviewed using prenatal ultrasound and/or multiplanar fetal MRI images. Results: Presented orbital pathologies are:hypo/hypertelorism, micro/anophthalmia, cataract, coloboma, optic nerve hypoplasia, persistent hyperplastic primary vitreous and orbital masses with discussion on the associated clinical syndromes. Conclusions: Early prenatal diagnosis of ocular malformations has an important role in appropriate genetic counseling and postnatal management.The identification of orbital anomalies can help in the search for associated cerebral or systemic anomalies.

The Purpose or Case Report: Advances in Neonatal Intensive Care have led to substantial improvement in survival of preterm infants of extremely low gestational age (ELGA) between 22+0 and 23+6 weeks gestation. ELGA newborns are more susceptible to several complications of prematurity. Recognition of the sonographic features of the normal brain and identification of potential short and long-term complications of these infants is paramount.High resolution transfontanellar ultrasonography is the baseline "gold standard" of care imaging modality to assess the integrity of the neonatal brain and potential complications encountered in ELGA newborns. Color and spectral Doppler US allows high reliability and precision in the evaluation of the intracranial vasculature.The purpose of this exhibit is1. To illustrate a spectrum of the sonographic features of the normal developing brain in ELGA infants.2. To characterize the sonographic findings of short and long-term brain anatomical complications.3. To describe common pitfalls when imaging these infants.

Prenatal Imaging Evaluation of Disorders of Sexual Development Tara Purpose or Case Report: Background: Disorders of sexual development (DSD) resulting in ambiguous genitalia are a rare spectrum of anomalies that have the potential to be diagnosed prenatally using a combination of genetic testing and imaging. The incidence of prenatal detection is rising with the increased use of noninvasive prenatal testing, which can reveal discordance between genotype and phenotype.Sonographic and MR imaging contribute to prenatal assessment of disorders of DSD and may narrow the differential diagnosis and facilitate prenatal testing and postnatal evaluation.The goals of this exhibit are:1. Review imaging features of normal prenatal male and female genitalia.2. Review imaging patterns of ambiguous genitalia.3. Discuss changes in appearance with various pathologies, providing imaging examples.4. Review different categories of disorders of sexual differentiation and see how imaging may narrow the differential diagnosis.

Fetal ultrasound and MR studies at our institute were reviewed retrospectively with selected representative cases chosen to illustrate technical aspects and to demonstrate the imaging features in cases of ambiguous genitalia. Correlation was made with NIPT, amniocentesis, and CVS results in conjunction with follow up radiology studies, and clinical or surgical outcomes.Categorical disorders of sexual development, such as 46 XX, 46 XY, ovotesticular DSD and sex chromosome will be described. Results: Prenatal findings in the normal fetus and in disorders of sexual development fetal will be illustrated with postnatal correlation. Conclusions: DSD is a complex group of disorders requiring an understanding of anatomical variations of genitalia and the diseases within each group. US and MR contribute to the prenatal detection of DSD, help narrow the differential diagnosis and help direct prenatal and postnatal genetic testing and imaging. Purpose or Case Report: The presacral space is composed of multiple tissue types, including osteochondral, mesenchymal, neurogenic, vascular and lymphatic. The presence of these tissues leads to a long and complex differential for a presacral mass in a pediatric patient. Specifically, the differential also includes anterior sacral meningocele, enteric cyst, vascular malformations, neuroblastoma, ganglioneuroma, schwannoma, neurofibroma, rhabdomyosarcoma, lymphomatous masses, giant cell tumor, aneurysmal bone cyst, osteosarcoma, Ewing sarcoma, and chordoma. Imaging plays a key role in characterizing these masses and treatment planning. Familiarity with the common presacral masses of infancy and childhood is therefore necessary for the pediatric and general radiologist.Once an osteochondral or neurogenic mass is excluded, and a predominantly cystic presacral mass is present, the primary differential consists of a collection of developmental or congenital masses including sacrococcygeal teratoma, anterior meningocele, low flow vascular malformation, and an enteric duplication cyst. We will review the common imaging features and associations of these cystic masses by presenting a series of cases. Additional examples of solid and osteochondral presacral masses will be included where appropriate for comparison in order to avoid characterization pitfalls and highlight teaching points. Purpose or Case Report: Prenatal screening and diagnosis of fetal anomalies rely upon ultrasound studies (US). US is generally considered safe during pregnancy. It is estimated that 1% of all neonates are born with a congenital heart defect that makes careful standardized examination of the fetal heart a mandatory step during all fetal US exams. US examination of the fetal heart requires at least analysis of the fetal situs, the four-chamber (4CH) view and the three-vessel view (3V). In the evaluation of the fetal heart, it is now worldwide accepted that the three-vessel trachea (3VT) view should be obtained, if technically feasible, in supplement of the other classical fetal cardiac views (4CH, 3V).The relevance of adding the 3VT view is that some congenital cardiovascular abnormalities, potentially significant for fetal outcome, can only be diagnosed with this view; these malformations include vascular rings and right aortic arches.

We conducted a retrospective review of prenatal US and cardiac US data of fetuses with abnormal 3VT view, with correlation with postnatal imaging (CTA) and outcome. The most frequent abnormalities detectable will be illustrated and discussed: right aortic arch with retroesophageal diverticulum of Kommerell, mirror-image right aortic arch, right aortic arch with aberrant left subclavian artery, right circumflex aorta, double aortic arch and left aortic arch with aberrant right subclavian artery.The goals of this exhibit are:-To review the normal appearance of the fetal 3VT view;-To review the embryological development of aortic arches explaining the malformations that can be observed;-To understand the significance of abnormal appearance of the fetal 3VT view by correlation with postnatal CT-scan imaging;-To propose a practical algorithm and work-up regarding detection of an abnormal 3VT view in a foetus;-To help radiologists provide appropriate counselling to parents when discovering an abnormal 3VT view in routine fetal screening. Purpose or Case Report: Congenital anomalies causing lower extremity shortening can result from dysgenesis or agenesis of the bones of the thigh, leg or foot; they are generally a very uncommon occurrence, with an incidence on the order of one case per 1,000,000 to one case per 100,000 births. These conditions result in varying degrees of morbidity, ranging from gait dysfunction to complete lack of the ability to ambulate. The accurate characterization of such an anomaly may be challenging prenatally, but can have significant impact on prognosis and treatment planning.We present a collection of cases depicting the spectrum of prenatally diagnosed anomalies of limb development, including proximal focal femoral deficiency, multiple cases of varying degrees of fibular and tibial hemimelia, amniotic band syndrome, benign uterine packing, neurofibromatosis, clubfoot anomaly and rocker bottom foot, all of which were evaluated on prenatal ultrasound and/or fetal MRI. Accurate prenatal diagnosis is extremely important for prognosis, treatment planning and riskstratification for associated congenital anomalies. The purpose of this poster will be to highlight the imaging features by both fetal MRI and prenatal US, discuss potential diagnostic pitfalls and review the clinical implications of this interesting spectrum of congenital disorders. We will describe a methodical approach to assessment of these patients. Our recommendations include: prenatal US to include a complete set of bilateral long bone and foot length measurements; lateral and footprint views of the fetal foot; views of both tibiae and fibulae bilaterally; views of the spine and upper extremities, a thorough search for other abnormalities; fetal MRI for complex cases or when US findings are limited; and a complete family and maternal history, including notation of family stature.Examples of the classic appearance of these conditions will be presented. Several missed diagnoses and the lessons learned will also be discussed. Counseling points addressed by the orthopedic surgeon will be included. Outcome management will be reviewed. Purpose or Case Report: This educational exhibit will demonstrate the strengths and weakness of Fetal MRI at 3 Tesla in the evaluation of multiple gestations.

Employing a case series of multiple gestation pregnancies evaluated via 3 Tesla MRI, this educational exhibit will highlight the strengths and weaknesses of imaging obtained at a higher magnetic field strength when compared the more routinely available 1.5 Tesla field strength.

Findings and pathology will be correlated with prenatal ultrasounds as well as post-natal imaging as applicable.

Results: Although the superior resolution obtained at higher field strengths can be advantageous in identifying and further characterizing fetal pathology, the sequences are more susceptible to artifact accentuation, particularly dielectric effect and motion.

The continued advancement of MR imaging technology has made higher field strength imaging available at a growing number of medical centers. Although imaging at 3 Tesla affords improved spatial resolution to visualize and define pathologic entities, the increased prevalence of artifacts at these higher field strengths, particularly in the setting of multiple gestations, requires familiarization on the part of both the technologist and radiologist to ensure diagnostic images are successfully obtained. Purpose or Case Report: Midface anomalies in the fetus can present with a range of severity and corresponding neonatal morbidity, on a spectrum from cosmetic disturbance to airway obstruction. Immediate postpartum complications include lifethreatening hypoxia and feeding disturbances. These anomalies not only present immediate difficulties but also often exist within a variety of syndromes with long-term consequences, affecting various organ systems. While screening ultrasound frequently can identify deviations from normal, fetal MRI may provide more detailed and high-resolution imaging for the characterization of midface anomalies and associated prenatal disease.The identification of facial anomalies in utero should prompt further investigation for associated abnormalities, and will also allow for improved prenatal counseling, which can prepare parents for the immediate postpartum management including surgical planning and resource allocation. It furthermore provides a foundation to shape family expectations and to begin to frame psychosocial support, allowing parents to engage mentally and emotionally with the medical and surgical course that awaits.In this presentation, we highlight the normal fetal midface as seen on MRI at various stages of development, and outline a systematic approach for evaluation of the fetal midface structures. Recognition of this typical appearance will allow the radiologist to identify the range of possible abnormalities that can occur and which can shape prognosis when properly identified. Various cases of midface anomalies will be reviewed and linked to their clinical significance, including cases with associated intracranial anomalies and deficiencies in swallowing and respiration.

Neuronal migrational anomalies and cortical malformations in Fetal MRI -Embryological perspective and avoiding pitfalls Shankar S. Ganapathy, MD 1 , sganapathy@akronchildrens.org, Kyle Hunter 2 , Emily Janitz 1 , Gayathri Sreedher 1 ; 1 Pediatric Radiology , Akron Children's Hospital , Akron, OH, 2 Aultman Hospital, Canton, OH Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Neuronal migration and cortical organization takes place primarily during the second trimester and early third trimester. Therefore, detection of cortical malformations and migrational anomalies on a fetal MRI in a maturing brain is difficult and may not be evident in some cases, if performed early.Familiarity with normal sulcation pattern in a fetus at various ages of gestation and a basic understanding of the embryogenesis of neuronal migration is essential to be able to make a diagnosis of these entities.Goals of the exhibit -Understand development of cerebral cortex and normal fetal MRI appearance at various stages of gestation.-Using this knowledge, be able to recognize various cortical malformations in a fetus and provide a comprehensive diagnosis regarding etiology, whenever feasible or when the appearance is typical or when there are ancillary findings-Avoid certain pitfalls and improve accuracy without overcalling or undercalling migrational anomalies in a fetal MRIOutline of the exhibitEmbryology of neuronal migration (during second and third trimesters)Normal sulcation pattern during mid and late trimestersEtiology of neuronal migration and cortical malformations-Multifactorial-Infectious-Genetic mutation/chromosomal-Ischemic insult-Metabolic disorders like ZellwegerAppearance of various migrational anomalies and cortical malformations in a fetal brainPolymicrogyriasubtle and extensive forms, appearance during various stages of gestation, correlation with post-natal imaging, typical appearance in certain genetic mutations, TORCH infections etcSchizencephaly -associations with absence of septum pellucidum and other abnormalitiesGray matter heterotopia -How to differentiate from subependymal nodules and hemorrhage?Lissencephaly -How to identify a true lissencephaly/pachygyria in a developing brain with immature sulcal pattern?Hemimegalencephaly, Microcephaly with simplified gyral pattern and other cortical malformations. Evolution of cortical malformations on imaging-Comparison between early and late gestation fetal MRIs in the same patient when available-Prenatal and post-natal MRI correlation of migrational anomalies, when availableAncillary CNS and extra-CNS abnormalities identified in fetuses with migrational anomaliesPitfalls and false positive/ false negative cases and how to avoid them. Purpose or Case Report: Nutritional deficiencies in children are prevalent both in the US and abroad often under-diagnosed unless symptomatic. Some nutritional deficiencies have conspicuous imaging findings that might help with diagnosis and/or follow up in a timely manner. Similarly, imaging might be needed to diagnose or follow complications related to malnutrition and eating disorders such as bulimia, anorexia nervosa, binge eating and pica. This educational exhibit will review radiographic, CT and MR imaging of nutritional deficiencies and complications of malnutrition and eating disorders in children Purpose or Case Report: Intestinal rotation abnormality (IRA) can lead to catastrophic events from midgut volvulus. In a child that presents with bilious vomiting, urgent surgery should be performed. Ladd's surgery is associated with complications in 10% to 20% of the patients. In this review we will show that in selected asymptomatic children with IRA, imaging can help decide if observation should be considered rather than surgery.

We reviewed upper gastrointestinal contrast studies, small bowel follow through contrast studies (SBFT), and colon enemas of children with a diagnosis of IRA who had surgery. We evaluated location of the duodenal jejunal junction (DJJ), position of the duodenum in the lateral view, and the anatomy of the colon. Location of the DJJ to the right of the left pedicle on the anterioposterior view or anterior in the lateral view was defined as IRA. We evaluated three types of IRA based on current literature: nonrotation (entire colon in the left abdomen, cecum in the left lower quadrant and the small bowel in the right abdomen), atypical malrotation (DJJ between midline and left pedicle), and typical malrotation (IRA that cannot be defined as nonrotation or atypical malrotation). Results: We will present an imaging algorithm for evaluation of malrotation and show imaging findings of all types of IRA as well as challenging cases. We will show cases with nonrotation when there was no need for surgery and other cases when Ladd bands were found. We will show cases of atypical malrotation. We found that if the duodenum is posterior and the entire colon is normal, the mesenteric root is broad and there is no need for surgery. We will also show indeterminate cases, when laparoscopy was necessary.We will show how anatomy of the colon helps to define the types of IRA. We will show cases where SBFT did not demonstrate the entire colon and was misleading. Enema should, therefore, be considered to evaluate the anatomy of the colon for definitive diagnosis.

We suggest considering a more stringent definition of atypical malrotation when there is also posterior position of the duodenum and normal colon. In asymptomatic children, especially at older ages, findings of nonrotation or atypical malrotation can provide an option for observation rather than surgery.

Eovist in Pediatric MR: What Can and Can't It Do? Narendra S. Shet, MD 1 , narendras@gmail.com; John F. Flynn, MD 1 , Ezekiel Maloney, MD 2 , Ramesh Iyer, MD 2 ; 1 Children's National Medical Center, Washington, DC, 2 Seattle Children's Hospital, Seattle, WA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: The purpose of this review is to provide an overview of applications of Eovist in the pediatric population, as well as pitfalls. We will begin with an introduction on the origins of Eovist and how it and other hepatocyte specific contrast agents (HSCAs) differ from conventional gadolinium based contrast agents (GBCAs). We will then provide a sample of imaging protocols when Eovist is used for hepatic mass evaluation and for biliary indications. We will review benefits of Eovist in characterizing a variety of pediatric hepatic neoplasms, including hepatoblastoma, hepatocellular carcinoma, focal nodular hyperplasia, hepatic adenomas, and regenerative nodules, with mention of pathologies that do not follow traditional patterns (i.e., inflammatory adenomas). After this, we will cover biliary applications, including choledochal cysts, biliary obstruction, and postoperative bile leak. Finally, we will review some limitations of Eovist; notably, we will mention several scenarios where the hepatobiliary phase could not be achieved, and recommendations for how to approach these cases. Purpose or Case Report: Abdominal pain is common presentation to the pediatric emergency department, and appendicitis remains the most common pediatric surgical emergency. While the imaging workup of pediatric appendicitis typically begins with ultrasound, MRI is increasingly being used for problem-solving and to further evaluate equivocal ultrasound results. With increased utilization, more and more mimics of appendicitis are being encountered on MRI. The purpose of this educational exhibit is to familiarize the practicing pediatric radiologist with these mimics through a case-based review in order to improve diagnostic accuracy and patient care.

We will present a series of common and uncommon mimics of appendicitis on MRI in the pediatric patient based on over 5 years of experience with limited pediatric protocol MRI for appendicitis. While the exhibit will focus on MRI findings, correlative US and CT findings will be included.

Results: This case-based educational exhibit will cover multiple mimics of appendicitis encountered in clinical practice, including but not limited to: omental infarct, Meckel's diverticulitis, pyelonephritis, ovarian torsion, tubo-ovarian abscess, isolated fallopian tube torsion, intussusception, and psoas myositis. Conclusions: Abdominal pain with concern for appendicitis is a common pediatric clinical presentation which relies heavily on imaging for diagnosis. The utilization of MRI for the evaluation of possible appendicitis continues to rise, and familiarity with common and uncommon mimics of appendicitis which may be encountered on MRI will help the practicing pediatric radiologist with interpretation of these studies and improve patient care. Purpose or Case Report: Epidermolysis bullosa (EB) is a group of rare genetic disorders that arise from at least 19 gene mutations for proteins that are involved in skin integrity. EB affects 1 out of 20,000 births in the United States and results in fragile skin that easily blisters from any minor friction or mechanical trauma. Other organ systems can also be seriously affected. The chronic skin inflammation and infections also places patients at risk for developing squamous cell carcinoma. Unfortunately, there is currently no cure for EB. The standard of care is supportive therapy and includes daily wound care, specialized dressings, and pain control.Because of the wide spectrum of systemic symptoms, there are numerous imaging findings that can be seen in patients with EB. These radiologic features can be categorized by body system including: 1) cardiac; 2) respiratory; 3) gastrointestinal; 4) genitourinary; 5) musculoskeletal; and 6) prenatal/fetal.There are various precautions that must be taken when performing any type of imaging study or anesthetic procedure for a patient with EB. These include but are not limited to: avoiding adhesives on the skin, providing special care when moving a patient, supporting pressure points on the imaging table, using large amounts of sterile water-based gel for ultrasound probes, and taking special care to protect the patient's skin, airway and oral cavity during anesthesia or sedation events.The purpose of this poster is to: 1) review the various clinical presentations, pathologies, and associated imaging findings involved in EB, 2) examine imaging and anesthetic concerns when dealing with patients with EB, and 3) discuss the imaging and anesthesia approach used when evaluating EB patients. Purpose or Case Report: In this exhibit we will describe a fairly new surgical procedure -total pancreatectomy and islet autotransplant (TPIAT). TPIAT is an uncommon procedure performed in children who suffer from chronic or recurrent acute pancreatitis. The purpose of the exhibit is to learn about the selection criteria, the surgery itself, and the imaging features of surgical complications. Candidates are selected for surgery based on a multitude of factors including severity and chronicity of pain, laboratory evaluation of pancreatic exocrine and endocrine function, and imaging features of chronic pancreatitis on MRCP and/or endoscopic ultrasound. In the surgical procedure, the pancreas and duodenum are resected, islet cells are isolated and then infused through the portal vein into the liver. A splenectomy is performed and continuity of the GI tract restored. Patients may spend three weeks in the hospital after surgery. Months after the infusion, the islet cells will be embedded within the liver and begin to produce insulin. The primary goal of this procedure is to relieve debilitating pain and improve quality of life, for which this procedure is around 90% successful. The secondary goal of the procedure is to retain islet cell function. In this regard, approximately 40% of the pediatric patients will regain complete islet cell function, 30% will regain partial function while 30% will have no islet cell function.

Follow up imaging may also be performed on these patients for evaluation of complications such as delayed gastric emptying, small bowel obstruction, surgical site infections, bile leak, and intra-abdominal hemorrhage. Complications seen in our hospital, including small bowel obstruction, delayed gastric emptying, and prolonged ileus will be highlighted. Purpose or Case Report: To retrospectively review the imaging planes performed, the number pyloric layers visible and the location of measurements taken, in infants with suspected (HPS).

103 pyloric ultrasound studies referred for suspected HPS were included. One pediatric radiologist with 20 years of experience and one medical doctor reviewed the studies. For each individual study, it was recorded whether longitudinal or transverse views were performed, the layers that could be visualised. A schematic was developed to categorise the interfaces of the anatomic layers of the pylorus visualised and position of the internal measurement cursor. Categories for the anterior wall were: a deep aspect of the muscularis propria; b superficial aspect of the muscularis mucosa; c deep aspect of the muscularis mucosa; d superficial aspect of the mucosa interfacing with the muscularis mucosa from a mucosal fold; e deep aspect of the muscularis mucosa from a mucosal fold. Categories for the posterior wall were: 1 deep aspect of the muscularis propria; 2 superficial aspect of the muscularis mucosa; 3 deep aspect of the muscularis mucosa; 4 deep aspect of the mucosa interfacing with muscularis mucosa from a mucosal fold; 5 deep aspect of the muscularis mucosa from a mucosal fold. Descriptive analysis was made for the categorical variables using STATA 15 software. Results: A total of 100 studies (97 patients) were reviewed. Studies recorded longitudinal (99%) and transverse (69%) views of the pylorus. For the longitudinal view, measurements included muscle thickness in 95%, length in 97% and 0% for the pyloric diameter. For the transverse view, measurements included muscle thickness in 16% and the diameter in 3%.

Pyloric layer interfaces (as defined above) were visible as follows: a in 64%, b in 64%, c in 66%, d in 30% and e in 26%. The internal (deep) reference point of cursor placement for measuring the muscle wall thickness in the longitudinal view was as follows: a in 46%, b in 27%, c in 30%, d in 1% and e in 2% of the studies. Conclusions: New US technology provides more detailed anatomy and this affects measurements for muscle wall thickness. Considering that a millimetre can make a measurement fall into the abnormal category resulting in surgical treatment such differences in practice must be highlighted and recommendations need to be clarified. We believe that the largely abandoned diameter measurement, in the transverse or longitudinal views, may offer a solution as it is not defined by any internal layers. Purpose or Case Report: Bowel obstruction in the neonate is common. When neonatal intestinal obstruction is suspected, initial workup may include abdominal radiographs or ultrasound; however, in most busy pediatric radiology practices fluoroscopy is indicated to diagnose the cause, which helps the surgeon make management decisions. Accurate diagnosis is key to the successful management of these neonates. The fluoroscopist should be mindful of the methods and techniques which make evaluation of distal bowel obstruction straightforward and efficient. This exhibit will detail the rationale and protocol to perform the optimal contrast enema, which is essential to have a chance to reliably distinguish the several causes of distal bowel obstruction in neonates. Will include examples with imaging findings and differential diagnoses. Purpose or Case Report: Direct impact upon the bicycle handlebars is an especially harmful mechanism of injury in children. Pediatric patients who present following direct impact of the bicycle handlebars upon the abdomen are much more likely to suffer internal organ injury and require operative intervention than those who present following a bicycle accident without direct impact upon the handlebars.Unfortunately, bicycle handlebar trauma is not only a damaging mechanism of injury, but also a treacherous one. Physical exam findings can be subtle; these children often appear misleadingly well. Many such children are initially misdiagnosed and sent home prematurely only to re-present. Delays in definitive diagnosis are typical.The radiologist, therefore, by aiding in the timely identification of injuries and determination of any need for surgery, can make a valuable contribution to patient care.Our cases include a 9-year-old girl with a puncture wound and abscess in the medial thigh musculature, a 13-year-old male with a penetrating wound to the left lower abdominal quadrant with extensive retroperitoneal and extraperitoneal hematoma and other injuries, a 14-year-old male with a hematoma in the right lower abdominal quadrant, and a 7-year-old boy with a liver laceration, each of whom suffered a bicycle handlebar impact to the abdomen.As companion cases, we also present a 12-year-old boy with a transected pancreas following an ATV rollover which involved impact of the handlebars upon the abdomen, and a 49-year-old man who sustained a shattered kidney in a bicycle accident.

Acute, emergent and post-surgical disorders involving the pediatric jejunum.

Deborah Brahee, Deborah.Brahee@cchmc.org; Alexander J. Towbin, MD; Cincinnati Children's Hospital Medical Center, Cincinnati, OH Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Pediatric pathology involving the jejunum is more common than one might initially expect. Early recognition of the important imaging characteristic, atypical findings, and useful imaging tools/techniques in the evaluation of jejunal pathology is important in prompt diagnosis and management of these patients.In this educational exhibit we will present a series of cases involving pathology of the jejunum encountered in the pediatric population. We will focus on important pathologies affecting a difficult to image and sometimes forgotten portion of the intestine.A range of acute, emergent and post-surgical cases are presented. Using an interactive, quiz based approach we will discuss the following pediatric pathologies that may affect the jejunum of pediatric patients from neonates to teenagers with multimodality imaging Purpose or Case Report: This poster aims to educate about the radiographic findings of pediatric umbilical disorders by presenting a case series, and to correlate the pathology with embryogenesis.

A retrospective review of our database was performed to select cases of umbilical disorders in children that required imaging. Imaging findings, pathological reports, and electronic medical records of these cases were reviewed. Results: Radiographic, surgical, and pathologic findings of 6 cases of pediatric umbilical disorders were reviewed from our database. Patient age ranged from 10 weeks to 11 years old. The cases included patent omphalomesenteric duct, infected urachal cyst, patent urachus, and Meckel's diverticulum. Conclusions: Umbilical disorders are classified according to embryonic remnants in the umbilicus and its associations with the urinary bladder and the bowel, which is why imaging can be useful in understanding the anatomy and differential diagnoses in such cases. These abnormalities can also have nonspecific clinical manifestations. Therefore, imaging can help in identifying and correctly diagnosing umbilical disorders, which is key to appropriate and timely patient management.

Pediatric Spleen: Anatomic Variants and Pathology James Davis, jdavi@umich.edu; Kathleen Gebarski, Peter Strouse; University of Michigan, Ann Arbor, MI Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: While the most common abnormalities of the spleen are splenomegaly and trauma, many additional disease states can be manifested in the spleen which should be considered when evaluating the abdomen on cross sectional imaging. We composed a pictorial guide of a wide variety of variants and diseases for education and reference.

Illustrative cases of congenital, developmental, infectious, and neoplastic processes involving the spleen were chosen from the last fifteen years of abdominal sonography at our institution along with CT, MRI and pathologic correlation. Case examples include polysplenia, storage diseases, wandering spleen, torsed spleen, various cysts, siderotic nodules, splenic complications of sickle cell disease including autolysis, islands of residual splenic tissue and echogenic foci, hamartoma, various infections, hemangioma, and infiltrative diseases such as leukemia and lymphoma. Results: Reviewing these cases provides an experience of a wide variety of splenic anatomic variants and pathology and improves the accuracy of interpretation.

A pictorial review of a wide variety of splenic anatomy and pathology improves the accuracy of interpretation. Purpose or Case Report: Pancreatic neoplasms are rare in children and young adults, with an incidence of 0.46 per million under 30 years of age. Fortunately, children with a pancreatic neoplasm have a better prognosis than adults. The better prognosis and rarity of disease both contribute to the fact that pancreatic malignancies account for less than 0.2% of cancerrelated deaths.Ultrasound is often the initial imaging modality to identify a pancreatic neoplasm due to its use for evaluation of symptoms such as upper abdominal pain, a palpable epigastric mass, or jaundice. Known pancreatic masses, or those detected initially by ultrasound should be imaged with CT or MRI to best characterize the mass and its relationship to adjacent structures, particularly the vasculature. Nuclear medicine plays a role in imaging of some pancreatic neoplasms depending on histology.Once a pancreatic neoplasm is identified, the radiologist is tasked with making a diagnosis from a differential diagnosis list of rare tumors. Primary pancreatic neoplasms are divided into epithelial and nonepithelial types. The epithelial tumors are more common and may be further subdivided into exocrine or endocrine subtypes. Epithelial exocrine tumors are the most common in children. Examples of these tumors include the two most common pediatric pancreatic neoplasms pancreatoblastoma and solid-pseudopapillary neoplasms. Endocrine tumors are uncommon. While functioning endocrine tumors can occur, non-functioning tumors are more common and are associated with syndromes such as von Hippel Lindau, multiple endocrine neoplasia type 1, and tuberous sclerosis. Nonepithelial tumors are also rare and include a number of different entities such as lymphoma, neurofibroma, and teratomas. Finally, the pancreas is an extremely rare site of metastasis. Pancreatic metastases can occur with multiple primary malignancies including neuroblastoma, rhabdomyosarcoma, and osteosarcoma.This exhibit will describe the imaging work-up of pancreatic tumors in children.

We will illustrate the different clinical manifestations and imaging appearances of the various pediatric pancreatic neoplasms. Purpose or Case Report: The spleen is rarely the first organ to come to mind in discussing pathology of the abdomen, and indeed may often be the last. However, there are a multitude of splenic processes and abnormalities that should be kept in mind when discussing the pediatric abdomen. Additionally, splenic abnormalities, or lack thereof, can also be clues to diagnosis in more difficult cases.

A retrospective analysis of multimodality imaging in pediatric patients demonstrating splenic imaging abnormalities, who presented to an urban children's hospital since 2005 is performed. Imaging and clinical history are correlated with laboratory findings where applicable. A variety of splenic abnormalities are selected for imaging review. Results: Review of splenic abnormalities is provided with imaging examples. Examples include metabolic, malignancy, trauma, and other disorders, including, but not limited to abnormalities in splenic size, cystic and solid splenic lesions, infiltrative/consumptive processes, and situs disorders. Conclusions: The spleen is often the forgotten organ of the pediatric abdomen, but there are many exciting things to be found in the left upper quadrant. Clinicians and pediatric radiologists alike should not disregard the spleen in abdominal cases, for there are many pathologies and clues to be found. Purpose or Case Report: To describe the spectrum of imaging abnormalities seen in eating disorders such as anorexia nervosa in the pediatric population.

Patients who were diagnosed with anorexia nervosa and had imaging studies for indications related to their eating disorder at our institution were included in this review. The imaging studies reviewed included chest, abdomen and pelvis radiographs, fluoroscopy studies of the upper gastrointestinal tract, studies, ultrasound studies of the abdomen, Computed Tomography of the brain, abdomen and pelvis and Magnetic resonance imaging studies of the brain. Results: A variety of imaging findings were identified in these patients including mild morphological findings to sequelae of multi-organ failure which will be illustrated with case examples on this review. Conclusions: Eating disorders can affect multiple organ systems with a spectrum of morbidity. Eating disorders manifest usually in teen age females and can be challenging to diagnose.

Awareness among pediatric radiologists about imaging findings in eating disorders is crucial to the diagnosis and management. Purpose or Case Report: Müllerian Duct Anomalies (MDA) are a broad category of congenital deformities of the urogenital structures due to abnormal development, fusion, or resorption of the fetal Müllerian Ducts. By definition, the Müllerian Ducts are the paired fetal structures that are located medially to the Wolffian Ducts and develop into the Fallopian Tubes, uterus, cervix, and upper two-thirds of the vagina. The multitude of symptoms associated with MDAs vary broadly in severity, ranging from patients with no symptoms and incidentally discovered MDAs to patients with primary infertility, recurrent pregnancy loss and other reproductive problems.There are well documented associations of MDAs with other genitourinary malformations. The most commonly reported is renal agenesis, which accounts for up to 30% of all associated renal anomalies in the setting of MDA, however abnormalities including ureteral remnants with ectopic insertion or collecting system duplication have been reported as well. The overall incidence of MDAs varies, but have been estimated to be approximately 1-5% within the general population and up to 13-25% in patients with recurrent pregnancy loss.The most commonly accepted system of classification of the various types of MDAs is that of the American Fertility Society, which divides MDAs into seven categories. However, there is often confusion as to how to categorize MDAs demonstrating characteristics of more than one class. Medical imaging is of essential importance for both this reason and for appropriate planning of therapy. The imaging of suspected MDAs involves the implementation of a combination of 3D Ultrasound, Magnetic Resonance Imaging, Sonohysterography, Hysterosalpingraphy and/or laparoscopy. Sonography and MRI have the benefits of being the least invasive methods of diagnosis.The goal of our educational exhibit will provide an overview of the embryology and pathophysiology related to a broad variety Müllerian Duct Anomalies. We will discuss the common imaging techniques employed in the workup of patients with Müllerian Anomalies. In addition, we provide a plethora of interesting cases from our home institution including a case of Mayer-Rokitansky-Küster-Hauser Syndrome, Obstructed Hemivagina and Ipsilateral Renal Anomaly (OHVIRA), cloacal abnormality, bicornuate uterus and other complex cases utilizing a variety of imaging modalities. Purpose or Case Report: Adnexal torsion is a common emergency room query in young girls and teenagers presenting with pelvic pain. It is a consequence of an underlying adnexal lesion or to anatomical laxity of the suspensory/anchoring ligaments. Clinical symptoms at presentation can be confusing and nonspecific, yet prompt diagnosis and surgical intervention are essential to save the adnexa. Radiology plays a crucial role in the diagnosis of adnexal torsion. Our exhibit will review and promote the rapid MRI as the first line, one-stop imaging in the diagnosis of adnexal torsion and other causes of pelvic pain. Some cases will be presented with initial ultrasound images and demonstrate how MRI increased the confidence in the diagnosis. All presented cases have pathology correlation and or operative reports and follow-up imaging. We will show various cases of ovarian and tubal torsions and their mimickers. In particular, we include torsions due to anatomical laxity of the suspensory ligaments, resulting from tubo-ovarian cysts and solid adnexal masses as well as cases of isolated tubal torsions. Table of Purpose or Case Report: Pelvic and scrotal pain are common emergent presenting symptoms in the pediatric population, and these patients are commonly imaged to evaluated for gonadal torsion. In this educational exhibit we will review the entire spectrum of gonadal and paragonadal torsions in the pediatric population, focusing on clinical presentations, key imaging findings, possible pitfalls in diagnosis, and next steps in management. It is essential for the pediatric radiologist to be familiar with this range of pathology in order to render an accurate and timely diagnosis, particularly given that some (but not all) of these diagnoses require emergent surgery.

Illustrative cases of gonadal and paragonadal torsions were selected from the past 5 years at our tertiary-case academic medical center with a dedicated pediatric hospital. Cases include, but are not limited to, the spectrum of ovarian torsion (from subtle findings of torsion in a previously normal ovary to cases of ovarian torsion with an underlying mass), the spectrum of testicular torsion (from a twisted spermatic cord with preserved testicular flow to more severe torsion with testicular infarction), isolated fallopian tube torsion, paratubal cyst torsion, and paratesticular appendage torsion. Results: A case-based review of pediatric pelvic (gonadal and paragonadal) torsions covering the entire range of presentation of common entities and classic presentations of rare entities will allow the pediatric radiologist to become more confident and competent in the interpretation of ultrasonography performed for these indications, and well as the interpretation of MRI performed for problem-solving. Conclusions: Suspected gonadal and paragonadal torsions are relatively common indications for imaging, particularly ultrasonography, in the pediatric population. Familiarity with these entities and the entire spectrum of imaging findings will improve diagnostic accuracy and efficiency of interpretation. This is extremely important as the pediatric radiologist plays a key role in making the ultimate diagnosis and distinguishing surgical emergencies from benign self-limited conditions.

Pediatric urodynamics: a radiologist's primer Emily Sellers, MD, emily.sellers@uchospitals.edu; Katherine L. Stahoviak, Seng Ong, MD, Kate A. Feinstein, MD; Radiology, University of Chicago, Chicago , IL Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity. Purpose or Case Report: In this exhibit, we will describe the synergistic effect of urodynamics and imaging. Bladder dysfunction, a common problem in children, accounts for up to 40 percent of pediatric urology clinic visits. Urodynamics is a key study to determine the etiology of bladder dysfunction yet, many radiologists are unfamiliar with this exam. Urodynamics is a fluoroscopic evaluation of the bladder with contrast during which abdominal and bladder pressures are measured while the bladder fills and empties. It uses a combination of electromyogram to evaluate pelvic floor contractions, cystometrogram to evaluate bladder pressures and capacity, and voiding cystourethrography to evaluate lower urinary tract anatomy. This exhibit will explain the indications for urodynamics, the basics of its interpretation including detrusor function and pressure flow studies, and the types of pathologies which may be diagnosed. Neurogenic bladder, congenital abnormalities, and how to differentiate between subsets of overactive bladder and underactive bladder will be explained. We will show imaging findings which correlate with these urodynamic diagnoses and describe ways to enhance voiding cystourethrograms and sonograms to better evaluate children with lower urinary tract dysfunction. Purpose or Case Report: Voiding cystourethrography (VCUG) has long been utilized in the diagnosis of vesicoureteral reflux in children with urinary tract infections. Sonographic techniques have been developed to make the same diagnosis without exposing the child to ionizing radiation. This presentation describes the implementation steps of a new service including: pharmacy approval and formulary addition; sonographer training and coordination with fluoroscopy technologists; referring clinician education and outreach. This exhibit educates the reader on how ceVUS can be performed, how it can be established at a children's hospital, and the benefits it provides.Correct technique and acquisition of images is shown using clinical case examples. Relevant anatomy is reviewed. Correlation of ultrasound abnormalities with VCUG results is provided. The process of training technologists and radiologists is highlighted. Pitfalls and tips to improve imaging are described. We will present results on how our new service has been accepted and utilized by the urologists and referring physicians. We have now performed over 50 ceVUS studies this year.

Poster Purpose or Case Report: Background: Image quality in MRI is often degraded by patient motion. To reduce the need for repeat exams, sedation or general anesthesia is used in pediatric patients. This requires additional clinical staff, and leads to increased scheduling wait times and overall procedure times that impact workflow. Many institutions prepare children for MRI using mock-up replicas. While effective in acclimating patients to the MRI environment, the availability of mock-up practices are often limited. Mock-ups also require physical space and do not simulate the full MRI environment and exam experience.Purpose: We build an affordable, scalable, and portable immersive virtual reality (IVR) platform for simulating MRI exams. Specifically, we designed an IVR environment that accurately mimics a Siemens 3 Tesla suite within our Radiology practice, including intricate details such as room size, color, lighting, ancillary equipment, pulse sequence audio, and background noise. The IVR platform also attempts to simulate the complete sequence of events and environments a child will experience during an exam, including the check-in process, interactions with staff (e.g., nurses, technologists), the waiting area and changing rooms, the positioning and motion of the MRI table, placement of coils, and within-exam instructions, such as breath-holds. Our platform also allows the patient to move around in the VR space and interact with the environment. Additionally, the platform can be easily converted to mimic any other MR suite. We hypothesize that our platform can be easily adopted by Child Life and hospital staff to quickly acclimate a patient and assess whether he/she can suitably undergo an MRI exam without sedation or general anesthesia.We believe the IVR platform can overcome the limitations of mock-up replicas. First, IVR has a smaller footprint and is scalable and portable across the hospital. This allows multiple IVR sessions to be held in parallel. Second, IVR can give the patient a hospital-specific and scan-specific first-person experience. The patient can interact with realistic 3D representations of the MRI environment and processes. Our current implementation of IVR was developed using HTC VIVE headsets with a dedicated laptop for control. To further improve portability, the IVR setup can be adopted using simple goggles and handheld devices. This will enable future IVR sessions to be conducted at the bedside, in patient's homes, and in referring physician offices.: Purpose or Case Report: Magnetic resonance imaging (MRI) is a robust imaging modality. However, the necessity for patients to hold still commonly requires pediatric patients to be sedated (which carries its own risks), or limits the useful scan time window in some nonsedated pediatric patients. In some circumstances, there is no way to control the movement of the imaging target (i.e. fetal MRI).MRI is also very customizable, and there are number of options currently available to accelerate MR imaging. It is important for pediatric radiologists supervising pediatric MRI examinations to have a broad understanding of the technology that is currently available, in order to optimize imaging quality and mitigate sedation use.In this educational exhibit, we will briefly review a number of acceleration techniques available on the Siemens platform, noting that other vendors have very similar options available for their customers. These techniques will include single-shot fast spin echo , balanced steady-state gradient echo, parallel imaging, radial imaging, and simultaneous multislice imaging. Our review will focus on the relative advantages and disadvantages of these techniques, rather than on the physics of image generation. Purpose or Case Report: Many diseases in pediatrics are relatively uncommon and imaging of these diseases is difficult to study due to small patient numbers at a single site. This limits the ability of single institutions to adequately power a study. Thus, there is a pressing need for a multi-site structure to combine data for rare diseases to appropriately power outcome studies. Our goal is to create an infrastructure to support pooling of imaging and clinical data across institutions to facilitate multi-institutional studies.

For multi-site studies, a centralized IRB structure is necessary to efficiently perform the study. The site establishing the centralized IRB will be considered the 'primary' site. The primary site is also responsible for IRB regulatory, data management, including imaging, and facilitation of project overall. The centralized IRB will establish reliance agreements with participating sites, enabling the sites to utilize the central IRB as the IRB of record. Once IRB approval is obtained, a centralized REDCap, or data entry system, at the primary site will grant access to participating institutions for data entry. Centralized data collection allows all data to be housed in one location, allowing for quicker analysis. For studies requiring imaging transfer, a standard operating procedure (SOP) for image de-identification, naming convention, and image transfer to centralized PACS system should be followed. Results: Our site has established three multi-site clinical imaging trials to date. First, we have an 11-site imaging repository for pleuropulmonary blastoma through the Midwest surgical consortium. All images have been uploaded to our site and pediatric radiologists at three different institutions are viewing de-identified images and inputting data in REDCap. Second, we have a three-site retrospective study evaluating the use of grayscale and Doppler imaging in veno-occlusive disease. Each site has access to centralized REDCap and data entry is performed on local subjects. Third, we have a two-site retrospective and prospective fetal MRI registry. Data collection and image upload to centralized PACS is being used to combine data between institutions.

We have successfully created infrastructure to support multi-institution clinical imaging trials. We hope that SPR members can use this resource for future studies on rare conditions where we can pool data to see how imaging affects patient outcomes. Purpose or Case Report: Fluoroscopy of the upper GI tract (UGI) can be difficult to master given the time-sensitive nature of the examination, the necessary hand-eye coordination, the complex button layout and broad featureset of the fluoroscopic equipment, the desire for minimizing radiation dose, and the required understanding of normal anatomy. Additionally, encountering abnormal findings for the first time, particularly before one is familiar with normal findings, can lead to confusion and increased difficulty in performing the examination. The use of 3D printed models of normal anatomy in pediatric fluoroscopic UGI training simulation has previously been demonstrated as a viable alternative to learning on live patients; however, such a technique has not previously been used with known pathological anatomic configurations.

A freely-available computer generated 3D model of the normal anatomy of the esophagus, stomach, and duodenum was downloaded and modified using Blender™ and Autodesk Fusion 360™. The model was edited to create 4 alternative models corresponding to the 4 classic patterns of abnormal UGI configuration: Malposition of the duodenaljejunal junction (DJJ), corkscrew duodenum, complete duodenal obstruction, and partial duodenal obstruction with normally positioned DJJ. Additionally, known normal variants of anatomy (such as duodenum inversum) were modeled as well.Inexpensive reusable simulator models were printed using a waterproof photo-reactive resin to allow residents to practice UGI fluoroscopy on both normal and abnormal patients.

Residents were tested on their performance and interpretation of UGI fluoroscopy on these models to enhance their comfort with performing the examination and to increase their familiarity with these important variants/abnormalities. Results: Though the models were entirely reliant on gravitydependent flow of contrast, they behaved similar to real neonates when filled with contrast and placed in different positions.

Conclusions: 3D printed models of the UGI tract can serve as adequate training tools for both general examination performance as well as education of normal anatomic and pathological variants. Purpose or Case Report: Successful bronchoscope handling requires the skill to orient bronchoscope position and direction in response to the intraluminal view provided by the bronchoscope camera. Additional challenges for pediatric physicians are smaller airways and the physiologically higher breathing frequency and airway collapsibility in babies and toddlers. We aimed to create a set of anatomically accurate 3D printed pediatric static and dynamic airway models that can be further used to teach and train residents/fellows in bronchoscopy and foreign body removal.

Three versions of 3D printed models were designed: a static tree model, a dynamic tree model, and translucent airway box model. CT patient data of three different ages (1, 5, 18) was selected for segmenting in Materialise Mimics. For the tree methods, the airway was then wrapped with a 2mm offset and hollowed out to create a lumen-like model and then was printed in the soft Tango+ material on Connex 500 or J750 printers. The branches were open for the static model and closed for the dynamic. The box was created by subtracting the airway from a box around its extents. This was then printed in VisoClear on a Project 6000HD printer. A y-connector and air sucking pump was attached to the dynamic model to simulate breathing and airway collapsibility.Three pediatric pulmonology attendings evaluated the models for physiologic accuracy and usefulness for teaching and training. Results: All models were evaluated to have an excellent intraluminal accuracy (branching and angles of bronchi, appearance of the lumen) and usefulness for teaching and training. The translucent box was favored for presenting and the static model to learn basic handling of the bronchoscope in bronchoscopy and foreign body removal. The dynamic model provided the most realistic cartilage consistency and endoscopic simulation of the physiologic breathing patient.Next steps planned are automatization of breathing simulation with different age-adjusted breathing frequencies and prospective enrollment of residents/fellows to formally assess technical skill development.

By fabricating three different 3D printed airway models, which enable visualization of bronchoscope handling and simulation of realistic intraluminal as well as physiological conditions in different age groups, we created a promising tool for teaching, training and testing pediatric residents/fellows in bronchoscopy and foreign body removal.

Poster Purpose or Case Report: It is important for radiologists to recognize and discuss with patients and families the potential risks and clinical manifestations of high Peak Skin Dose. In children undergoing complex Interventional procedures accurate Peak Skin Dose estimates are complicated and time consuming despite reference point doses being available. The availability of the Dose Structure Report (SR) on modern fluoroscopic equipment allows such estimates to be timely and consistent. The process of creating a Peak Skin Dose estimate using a validated radiation data management system (RDMS), capable of collecting detailed acquisition data and modeling will be discussed.

Interventional fluoroscopic equipment is configured to transfer the SR to the hospital PACS system (PACS). PACS sends SR to a RDMS, which interprets SR reports and provides details of each study acquisition including: dose area product, reference point dose, beam on time, pulse rate, pulse width and height, table positions, collimation, preliminary and secondary angulations, distances of source to table and receptor. Acquisition protocols, timeline, and angulation maps are generated for each study. A Peak Skin Dose estimate for each study is performed using the detailed acquisition data, with manual adjustments made for table offsets in lateral and longitudinal directions, patient orientation and skin model (i.e. patient size). The skin dose map can be captured and used for initial and follow-up discussions with patients and families.

Results: For all studies, a default Peak Skin Dose estimate and a color skin dose map are automatically generated. With a pre-set institutional threshold (1 Gy) for Peak Skin Dose evaluation, RSMS automatically alerts the lead staff members of IR and physicist if the threshold level is exceeded upon completion of the study. The Peak Skin Dose estimate is then re-calculated with minor adjustments as described above. The skin dose map can be captured and used for initial and follow-up evaluations with patients and families.Two cases are selected for demonstration: 1) >1Gy Reference Point Dose, < 2Gy Peak Skin Dose; 2) >1Gy Reference Point Dose, > 2Gy Peak Skin Dose.

Conclusions: Using SR in RDMS for estimates of Peak Skin Dose can provide clinicians with timely Peak Skin Dose data to inform clinical patient follow up. This technology can also be a useful tool in the investigation of a potential sentinel event and in planning future similar interventions Purpose or Case Report: 1) Reviewing the pathology and outcomes of children with spinal muscular atrophy (SMA)2) Learning the procedures available for intrathecal access3) Understanding the indications for choosing more advanced techniques to deliver nusinersenSpinal muscular atrophy is an autosomal recessive disease affecting motor neurons and is the most common genetic cause of death in infants. Nuinersen (Spinraza) was recently approved by FDA for intrathecal administration in SMA patients. Commonly administered by lumbar puncture in the clinic or with imaging guidance, this is the initial method of administering nusinersen. Deformities and spinal instrumentation from orthopedic surgeries are common in SMA patients, preventing traditional intrathecal access by lumbar puncture for nusinersen delivery. Transforaminal lumbar sac access, ultrasound or fluoroscopy guided cervical spine access, or subcutaneous catheter placement can be the alternative approaches with failed/difficult lumbar access. Given the potential benefit of nusinersen, understanding all methods to obtain intrathecal access is essential for a pediatric interventional radiology (IR) practice.

We intend to review the technique, indications, complications, tips and tricks for each of these techniques and their relevance to an IR practice.1) Ultrasound guided lumbar puncture2) Fluoroscopy guided lumbar puncture3) Ultrasound guided cervical puncture4) Fluoroscopy guided cervical puncture (C1-C2)5) Transforaminal delivery using cone-beam computed tomography6) Transosseous access via drill7) Subcutaneous Intrathecal Catheter System Conclusions: Nusinersen administration for SMA is providing a novel treatment for a previously untreatable condition and the number of patients requiring these procedures is expected to increase. Pediatric interventional radiologists can provide the safest approach for delivering nusinersen by being aware of the options. Purpose or Case Report: To familiarize the audience with our institution's technique utilizing ultrasound (US) and electrocardiogram (ECG) for anatomic localization and technical guidance for PICC placement in a pediatric intensive care unit (PICU) procedure suite utilizing intravenous (IV) sedation without general anesthesia to eliminate patient and operator radiation exposure and decrease sedation specifically in a pediatric population.

Purpose or Case Report: Objectives:1. Present the use of a published algorithm for the evaluation and diagnosis of the pediatric patient with congenital skeletal dysplasia and abnormal skeletal survey.2. Review usefulness of accurate, narrowed differential diagnosis or suspected single diagnosis in terms of confirmatory testing, treatment implications, and genetic counseling.3. Demonstrate the utility of the algorithm when applied to recent, rare cases of congenital skeletal dysplasia at our institution, a tertiary trauma center and children's hospital in the Northeastern United

States.Content:We present a refined, algorithm-based approach to the evaluation and diagnosis of the pediatric patient with congenital skeletal dysplasia and abnormal skeletal survey. The algorithm optimizes evaluation of the skeletal survey in cases of congenital skeletal dysplasia, aiding in timely, accurate diagnosis. The utility of the refined algorithm is demonstrated as it was applied to recent, confirmed cases of rare skeletal dysplasias at our institution, including metatropic dysplasia and cleidocranial dysplasia.Teaching Message:Evaluation of the pediatric patient with congenital skeletal dysplasia and abnormal skeletal survey can be challenging, even for the subspecialty-trained radiologist. By assessing the presence or absence of discriminating imaging features and findings on skeletal survey, the interpreting radiologist can significantly shorten the differential diagnosis or in many cases suggest a single, most-likely primary diagnosis. Narrowing the differential diagnosis is helpful in guiding confirmatory molecular or genetic testing. Timely, accurate diagnosis may have significant treatment and prognostic implications for patients and their families.

When soft tissues turn hard -ossifying soft tissue lesions. Purpose or Case Report: The purpose of this educational exhibit is to illustrate the radiological appearance of various ossified soft tissue lesions. The differential diagnosis is wide, with entities ranging from benign to malignant. Using instructive and interesting cases encountered at our department, this presentation will outline an approach to forming a differential diagnosis. Various teaching points will be highlighted. Pathological correlation will be included where available.The cases to be presented include:Myositis ossificansFibrodysplasia ossificans progressivaTenosynovial osteochondromatosisDysplasia Epiphysealis HemimelicaPseudohypoparathyroidism -Albright's hereditary osteodystrophyPhlebolith in a vascular malformationPilomatrixomaHeterotopic ossificationSynovial SarcomaThis presentation will highlight the distinguishing imaging characteristics of each entity and help narrow the differential diagnosis for a radiologist when faced with a pediatric ossified soft tissue lesion. Purpose or Case Report: At our institution and in general practice, we have observed that ultrasound is an underutilized modality in the field of pediatric musculoskeletal imaging, particularly when compared to its body imaging counterparts and the adult population. Within the spectrum of musculoskeletal radiology, ultrasound can be an especially helpful tool in evaluating the peripheral nervous system (PNS), which can often be challenging to assess on other imaging modalities. For instance, ultrasound offers the ability to carefully trace small nerves and easily allows for comparison to the other, often unaffected or normal extremity, as well as dynamic imaging. This advantage is often not possible on magnetic resonance imaging (MRI), which is currently the primary modality being utilized to assess for nerve pathology.

In small children, nerves are often difficult to visualize on MRI, which can be limited by artifact and motion and may require sedation. Due to its superior axial resolution, ultrasound can better visualize the intra-neural architecture, sometimes better than MRI, especially in the case of small nerves. Also, as was demonstrated by one of the cases at our institution, ultrasound was superior in evaluation when there was adjacent surgical hardware which resulted in artifact on MRI. Additionally, we as radiologists, can aid our peripheral nerve surgeon colleagues intra-operatively with ultrasound.Some of the hesitance in using the modality may be a reflection of a lack of familiarity on the part of both radiologists and technologists with respect on how to image the PNS and uncertainty about the normal and abnormal appearance of these structures. The objective of this educational exhibit is to provide the knowledge necessary to successfully acquire and interpret ultrasound images of the PNS. The topics addressed in the review will include an introduction on how to appropriately obtain images of the major peripheral nerves, how to distinguish between the normal and abnormal appearance of nerves on ultrasound, and the spectrum of pathology affecting the peripheral nervous system. The exhibit will feature a variety of cases from our institution, including examples of neuropathy, post-traumatic focal neuroma, nerve laceration resulting in transection, ulnar nerve subluxation/dislocation, and nerve sheath tumors, among others, with some intraoperative and pathologic correlation.

Clinical update on high resolution magnetic resonance imaging of small joints in children and adolescents using 3D proton density fat suppressed turbo spin echo imaging accelerated with compressed sensing Purpose or Case Report: Ultrasound is typically the first-line imaging modality for the evaluation of superficial soft tissue masses in the pediatric population. While certain superficial soft tissue masses have a classic appearance on ultrasound, others may have a nonspecific appearance and may vary in their sonographic characteristics. This can make a definitive diagnosis and follow-up recommendations difficult, especially for trainees.We aim to provide a basic overview of some of the common and less common superficial soft tissue masses that may be encountered in pediatric imaging, and introduce a novel, structured algorithmic approach for evaluating these lesions on ultrasound. The algorithm will assist the radiologist in reaching a definitive diagnosis or narrowing the differential such that a helpful recommendation for further workup can be made. For example, the algorithm will include internal vascularity, border distinctness, cystic or solid components, presence or absence of calcifications, and location in the body and within the superficial soft tissue layers. Pictorial examples of each sonographic feature in the algorithm will provide further assistance.Pathologies will include but not be limited to: pilomatricoma, glanuloma annulare, epidermal inclusion cyst, ganglion cyst, abscess/infection, lipoma, hematoma, lymph nodes, vascular anomalies, and soft tissue sarcomas. When a definitive diagnosis is not achievable, this algorithm will help the radiologist determine the likelihood of benignity, a short differential diagnosis, and a recommendation for any further imaging workup.

Imaging Purpose or Case Report: Gaucher disease is an inherited metabolic disorder due deficiency of the lysosomal enzyme βglucocerebrosidase that results in the accumulation of abnormal macrophages ("Gaucher cells") within multiple organs, most conspicuously affecting the liver, spleen and bone marrow. As the diagnosis is increasingly made during childhood and young adulthood, pediatric radiologists should be familiar with imaging features of Gaucher disease and its complications.Visceromegaly consisting of hepatosplenomegaly is a hallmark of Gaucher disease and uniformly present in cohorts of pediatric patients. In addition, bone marrow involvement with 'Erlenmeyer flasks' have been long recognized as part of this disease, although the classic radiographic finding is not present until adulthood. Marrow involvement confers significant morbidity for these patients with pain, bone infarcts and pathologic fracture.Traditionally, imaging of disease severity has been based on hepatic and splenic visceral organ enlargement and/or qualitative assessment of bone marrow involvement. However, advances in the understanding of Gaucher disease and observations of elevated ferritin levels and increased risk of hepatic fibrosis emphasize the importance of more comprehensive assessment of liver involvement beyond simple enlargement. Moreover, quantitative MRI assessment of bone fat-fractions also may have a role in assessing marrow involvement. These methods of disease assessment are important in addressing management decisions regarding enzyme replacement and substrate reduction therapy.

Conclusions: This presentation summarizes the imaging evaluation of Gaucher disease as it pertains to pediatric patients. We review multimodal conventional imaging manifestations of Gaucher disease from radiographic manifestations to MRI appearances. Semi-quantitative marrow scoring methods are discussed. This exhibit also discusses newer quantitative approaches to assessment of liver and bone marrow involvement with an emphasis on future applications of advanced methods including spectroscopy, elastography, fatfraction and iron quantification in guiding therapy decisions and monitoring treatment response. Purpose or Case Report: A term "ciliopathy" represents a diverse group of genetic disorders caused by mutations in genes coding for components of the primary cilia. Primary cilia have a pivotal biological role in the cell surface of nearly every organ system of the body. "Skeletal ciliopathy" is a subset of ciliopathy mainly affecting the skeleton and shares common radiological findings such as short ribs, short limbs, and short digits with or without polydactyly. Pattern recognition approach is thus useful to diagnose skeletal ciliopathy. Skeletal ciliopathy includes (1) Jeune asphyxiating thoracic dysplasia, (2) Ellis-van Creveld syndrome (chondroectodermal dysplasia), (3) Sensenbrenner syndrome (cranioectodermal dysplasia), and (4) Short rib-polydactyly syndromes. Clinically, affected patients commonly present with thoracic hypoplasia with respiratory failure and disproportional stature with a normal trunk and short limbs most severe in the distal segments. Brachydactyly is conspicuous. Patients may have extra-skeletal anomalies such as retinopathy, cardiac anomalies, cerebellar malformations, and hepatorenal failures.Radiological diagnosis of bone dysplasia might be regarded as something complex. We aim to highlight a pattern recognition approach to diagnose skeletal ciliopathies, one of the major bone dysplasias, by providing many cases. We will also review a general concept of "bone dysplasia family" which refers to a grouping of radiologically similar skeletal disorders into a "family." This concept has been widely accepted now after we found that phenotypic similarities usually indicate the same/similar pathogenetic mechanisms, and it supports the use of a pattern recognition approach. Imaging diagnosis can guide genetic testing, interpretation, and possibly identify new genes or mutations.

In-phase and opposed-phase evaluation of bone marrow lesions in the pediatric population. Purpose or Case Report: Bone marrow is composed of fat and cellular elements supported by the trabecular bone. Pathological lesions of the bone marrow usually replace its normal constituent to a variable degree. Differentiation among traumatic, neoplastic, and inflammatory processes of the bone marrow is often not possible with MRI, which commonly requires the patient to undergo additional invasive diagnostic procedures to obtain an accurate diagnosis.In-phase and opposed-phase imaging has been used extensively in imaging of the liver and adrenal glands. However, recently it has been introduced into the evaluation of the bone marrow. The technique takes advantage of different excitation frequencies of water and fat protons due to differences in their molecular environments. The main concept is that the presence of normal bone marrow would result in suppression of signal intensity on Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Proximal humeral epiphyseal fracture-separation is a rare fracture pattern in infants often associated with birth-related or non-accidental trauma, representing a Salter-Harris type I or type II fracture. Lack of a proximal humeral epiphyseal ossification center in most newborns or only subtle displacement of a small epiphyseal ossification center in older infants, makes this injury difficult to diagnose on plain radiographs, potentially leading to delayed diagnosis or misdiagnosis. Ultrasound and MRI are therefore useful imaging modalities in indeterminate cases.Clinical findings of infant proximal humeral epiphyseal fractureseparation such as shoulder swelling, tenderness, and decreased active motion, overlap with more common entities including clavicular fracture, brachial plexus injury and osteomyelitis. Radiographs are often the first diagnostic study ordered to evaluate these symptoms but are insensitive due to minimal ossification of the proximal humeral epiphysis. Radiographs may be normal or show subtle displacement of the epiphyseal ossification center, apparent joint space widening, small metaphyseal fracture fragments or displacement of the proximal humeral metaphysis in relation to the scapula. These findings can be misdiagnosed as shoulder dislocation or pseudosubluxation due to a joint effusion. Careful review of the medical record may elucidate a history of difficult delivery with shoulder dystocia or suspected shoulder trauma.Ultrasound and MRI are useful in indeterminate cases due to their ability to resolve the cartilaginous physis and proximal humeral epiphysis, and to resolve their relationship to the humeral shaft and cartilaginous labrum. Ultrasound has higher anatomic resolution and offers the flexibility to quickly image the asymptomatic contralateral shoulder and image in planes that best show the relationship between the non-ossified epiphysis and the humeral shaft. Doppler ultrasound demonstrates epiphyseal perfusion without the need for contrast administration, and serial ultrasound imaging can be used to evaluate healing and remodeling. MRI is more useful in evaluating cases where osteomyelitis and/or septic arthritis are being considered, or in cases of an inconclusive history suspicious for non-accidental trauma to evaluate for additional osseous and soft tissue injuries.Several examples of typical proximal humeral epiphyseal fracture-separations in infants will be presented and the relevant imaging findings discussed. Purpose or Case Report: Background:AVID is an acronym describing a triad of findings including 1) asymmetric ventriculomegaly, 2) interhemispheric cyst, and 3) dysgenesis of the corpus callosum. This entity accounts for one of the presentations of callosal dysgenesis along a wide spectrum.

Because midline anomalies occur with many processes, including holoprosencephaly and aqueductal stenosis, it may appear to have overlapping features on initial glance. However, by focusing attention on the secondary findings, a specific diagnosis may be determined.Objectives:By the end of this presentation the learner will: 1) Become familiar with the imaging characteristics of AVID. 2) Describe the differential diagnosis of AVID and the important distinguishing features. 3) Understand the clinical implications of interhemispheric cysts and similar diagnoses.Cases/Differential Diagnosis:Through several case examples of mistaken diagnoses, we will describe key findings to differentiate brain disorders with midline anomalies including AVID, holoprosencephaly, and aqueductal stenosis. Holoprosencephaly creates a monoventricle, but may also be associated with a dorsal midline cyst which can be confused for an interhemispheric cyst. Features that differentiate holoprosencephaly from AVID are the presence of fused cerebral hemispheres, thalamic fusion, and a true monoventricle. Aqueductal stenosis may also show severe hydrocephalus, but lacks the cystic component which is seen with the other mentioned entities. Aqueduct stenosis usually causes symmetric ventriculomegaly of the lateral and 3rd ventricles as well as upward displacement of anterior cerebral arteries and inferior displacement of internal cerebral veins. The hydrocephalus from all of these entities may be treated with ventricular shunt placement, but AVID is an important diagnosis to consider because the wall of the interhemispheric cyst could be imperceptible by imaging and may not improve if the tip of the drainage catheter is not within the cyst. Conclusion: When evaluating cases of true ventriculomegaly it is important to consider a differential including AVID, holoprosencephaly, aqueductal stenosis, among other less common congenital syndromes. Careful attention to additional imaging findings is necessary to distinguish the correct diagnosis from look-alikes. Making an accurate diagnosis is important as there are differences in medical decision making, treatment outcomes, and long-term prognosis.

Acquired non-traumatic temporal bone lesions in children: A Pictorial Review.

Ashok Mithra Karuppiah Viswanathan, ashokmithrakv@gmail.com; Nagwa Wilson; CHEO, University of Ottawa, Ottawa, Ontario, Canada

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: A number of acquired non-traumatic diseases of myriad aetiologies involve the temporal bone in children. While some of these are also noted in adults, many diseases are specific to the pediatric age group. These can be grouped into infectious/inflammatory, neoplastic, vascular and other miscellaneous disorders.Anatomy of the temporal bone is complex. It forms the lateral aspect of skull base and comprises of five osseous parts viz. squamous, mastoid, petrous, tympanic and styloid segments. Specific disease processes afflict each part of the temporal bone, largely dictated by its anatomy and constituent structures. Hence a structured approach to image interpretation and reporting is especially useful in this region to localise the lesion and subsequently generate differential diagnoses.Traditionally CT has been the imaging modality of choice in assessing temporal bone lesions. However, currently, CT and MRI are deemed complimentary. CT provides exquisite details of anatomy, characterises osseous lesions, determines bony involvement/destruction and extension while MRI is highly useful in assessing intrinsic lesion characteristics due to its superior contrast resolution. In certain aetiologies such as cholesteatoma MRI is diagnostic.This poster aims to review the anatomy of temporal bone and various common, uncommon acquired non traumatic temporal bone lesions in children.The lesions that will be discussed in this poster are listed below:Infectious Purpose or Case Report: PET/MRI has recently become a clinical realization after overcoming complex hardware and image reconstruction issues. The goal of this educational exhibit is to provide a comprehensive, yet understandable, introduction to these aspects of PET/MRI along with displaying a pictorial assay of different normal and abnormal metabolic findings within the field of Pediatric Neuroradiology.The first part of this presentation will begin by highlighting the basic hardware components of the PET/CT contrasting with the interactions between the main components of the PET/MRI scanner along with their associated solutions. In general, these issues include how MRI can affect PET in terms of their magnetic field and RF properties and how PET affect MRI due to the scintillator/electronic components.The second part will begin by discussing some soft tissue and hardware attenuation correction techniques that are currently in use, such as: Segmentation and atlas-based methods along with attenuation map generation and coil localization methods. Additionally, we will show the consequence of field-of-view (FOV) mismatch between the PET and MRI acquisitions and partial volume effects along with their solutions.The final part will showcase clinical applications of PET/MRI to Pediatric Neuroradiology, featuring imaging protocol details and a pictorial guide of normal distributions and pathologic conditions. Clinical examples range from seizure localization, cortical malformations, manifestations of Phakomatoses, perinatal stroke, tumor recurrence, and Flutriciclamide ( 18 F-GE180) imaging in the setting of neuroinflammation. Purpose or Case Report: Dermoid cysts are a sequestration of ectoderm containing secondary skin structures, usually at characteristic locations of embryologic fusion lines. The midventral suprasternal fusion line is one such location in the developing embryo with both simple and complicated dermoids of this region reported in the literature. We reviewed our academic pediatric institutional experience with dermoid cysts of the suprasternal notch, focusing on the initial presentation as a palpable nodule referred for sonographic evaluation.The purpose of this educational exhibit is to depict the ultrasound imaging features of suprasternal notch dermoid cysts in the pediatric population.Secondary purposes are to describe the anatomy of the suprasternal notch, demonstrate CT and MR correlates of the these dermoids, and depict regional inflammatory complications.

Methods & Materials: 47 retrospective cases aged from 2 months to 13 years were collected utilizing an electronic database of radiology reports at our academic children's hospital from January 2008 to July 2018. Many cases were confirmed either with typical MR appearance or excisional biopsy.

Results: Suprasternal notch dermoid cysts demonstrate specific sonographic findings:(1) circumscribed ovoid mass located between the sternornohyoid muscles in the midline(2) longitudinal orientation(3) homogeneously echogenic with some of them showing small internal anechoic areas(4) posterior acoustic enhancement(5) absence of internal Doppler blood flow Conclusions: Characteristic location of a nodule in the suprasternal notch and specific sonographic findings allow for a confident radiologic diagnosis of a dermoid cyst without the need for additional imaging. Purpose or Case Report: The goal of this educational exhibit is to use a graphical and image heavy Powerpoint presentation to familiarize the reader with the background necessary to understand common clinical ASL-PWI imaging techniques along with displaying a pictorial assay of different normal and abnormal ASL perfusion findings within the field of Pediatric Neuroradiology.To introduce our topic, we will briefly discuss the physiology of cerebral blood flow and how it is measured with and without an exogenous tracer. Next, to serve as an overview, we will show the general ASL experiment, namely, labeling, post label delay, and readout. Following, we will discuss each of the above components, beginning with showing the different main labeling methods, CASL, PCASL, and PASL. Finally, the concept of post label delay will be illustrated respective to each of the main labeling methods.In the second section, we will discuss the image acquisition component by first describing and illustrating two common readout methods, such as 2D-EPI and 3D-GRASE methods along with their advantages and disadvantages. Additionally, we will illustrate the process of background suppression along with its significance. Finally, we will review the process of obtaining perfusion weighted images through the subtraction between tagged and control images.The third section will illustrate an assortment of clinical examples beginning with showing normal physiological hyper-and hypoperfusion and newborn perfusion characteristics. Following, we will highlight a spectrum of clinical cases including: moyamoya, HIE, medullary infarct with luxury perfusion, arteriovenous malformation, PRES, infections (intracerebral abscess, toxoplasmosis, cerebellitis), characteristics of headache, intracranial tumors (juvenile astrocytoma, hemangioblastoma, ATRT, and choroid plexus papilloma), cortical dysplasia, tuberous sclerosis with seizures, and variations presentations of methotrexate toxicity. Lastly, we will discuss cerebral blood reserve imaging utilizing acetazolamide challenge. All cases will contain pertinent clinical information and images from other sequences/modalities that aid in the diagnosis of disease. Purpose or Case Report: Epilepsy can be a progressive and debilitating illness in the pediatric population. There is a large range of congenital conditions that present with seizures in neonatal patients, which are essential for the radiologist to accurately characterize on imaging. Accurately diagnosing the cause of epilepsy can not only allow appropriate treatment, but also provide important information on prognosis and associated abnormalities.The goal of this educational poster is to provide an overview of different congenital etiologies of epilepsy. This case based review includes congenital malformations of the brain such as focal cortical dysplasia, schizencephaly, heterotopias, hemimegalencephaly, and polymicrogyria. Cases highlighting neurocutaneous conditions associated with epilepsy including tuberous sclerosis and Sturge-Weber are reviewed.

Cortical injuries from insults such as congenital infection, hypoxic-ischemic injury, or hemorrhage are also included.Familiarity with the imaging findings of epilepsy is essential for the radiologist, especially the trainee and those with less experience with pediatric neuroimaging. After reviewing this educational poster, viewers will be able to describe the imaging features of congenital causes of epilepsy and understand the treatment and prognoses of these conditions.

Imaging of Pediatric olfactory system anomalies The technique used to acquire images and the protocol used are described. Velocities, direction of flow, and waveforms are discussed. The template used to report the results is reviewed. The relationship of the major vessels to the probe is depicted graphically. Recommendations in the use of this technique to determine brain death are noted. Results: A 3V sector probe is used with transtemporal and transforaminal approach to interrogate the Circle of Willis including the middle cerebral artery, anterior cerebral artery, posterior cerebral artery, distal internal carotid artery, basilar artery and vertebral arteries. Mean velocities and pulsatility indices are recorded. Detection and evaluation of the hemodynamic effects of severe stenosis or occlusion of the extracranial (greater than or equal to 60% diameter reduction) and major basal intracranial arteries (greater than or equal to 50% diameter reduction) are possible. Transcranial Doppler has the advantage of providing real-time evaluation of cerebral vasculature without the use of contrast agents, ionizing radiation, or sedation. Conclusions: Imaging transcranial Doppler can be a useful diagnostic aid in the approach to managing treatment in the pediatric neurocritical care unit.

Ultrasonographic approach of neck masses in the pediatric population. Purpose or Case Report: Neck masses are commonly seen in the pediatric population and are a frequent reason for pediatric consults and can represent a diagnostic challenge. Although there is a broad spectrum of differentials, the vast majority of these lesions are benign, however malignant etiologies can also be encountered. Ultrasonography has become the first-line imaging modality in the evaluation of neck masses, given the ability to avoid radiation, wide availability and costeffectiveness. Also, ultrasound represents an option when trying to avoid contrast administration or sedation.The main teaching points of this exhibit will be:1. Purpose or Case Report: Poland Syndrome is a classic differential consideration for a unilateral hyperlucent hemithorax on chest radiograph due to pectoralis aplasia or hypoplasia. Additional associations include brachysyndactyly, simian crease, dextrocardia, and liver/biliary abnormalities. Isolated pectoral hypoplasia or aplasia without limb involvement is generally cosmetic and can be corrected with plastic surgery, but Mobius syndrome should be ruled out due to its association. Children with Mobius syndrome can exhibit an expressionless affect, excessive drooling, and/or eye paralysis due to cranial nerve deficiencies. This educational exhibit discusses the pediatric radiologist's role in the imaging approach to these supposedly separate diseases which may represent a continuum of one pathology.

The epidemiological, clinical, genetic, and imaging findings of both Poland syndrome and Mobius syndrome are described. A discussion of the appropriate protocol required to aid in the diagnosis of Mobius syndrome is included. Differential diagnoses are explored. Radiologic examples of additional associations such as morning glory syndrome and Pierre-Robin syndrome are also depicted. Results: Poland syndrome can occur sporadically, but some familial associations have been observed. It is more common in males and more commonly affects the right side of the body. Mobius does not appear to exhibit a gender predominance. Like Poland syndrome, the disease commonly occurs sporadically, though associated genetic mutations have been characterized. Imaging findings include cranial nerve hypoplasia or aplasia. Due to the association between these diseases, the name Poland-Mobius syndrome is sometimes uses. In the child presenting with Poland syndrome, MRI of the brain with high resolution imaging of the cranial nerves is indicated. Purpose or Case Report: Brainstem anomalies pose a diagnostic challenge for trainees to even the most advanced in their practice. Our exhibit aims to provide a comprehensive review of these rare disorders which are being diagnosed earlier and more frequently utilizing advanced imaging and genetic sequencing techniques. Representative images of these disorders with key distinguishing features will be presented. Familiarity with these conditions will benefit practicing radiologists, radiologists in training, and our clinical colleagues.

A diagnostic imaging approach to brainstem anomalies using high quality representative images with emphasis on MRI, including fetal MRI, will be presented. Results: 1) Review the appropriate stages of brainstem development.2) Highlight a number of cases demonstrating brainstem anomalies including tubulinopathies, dystroglycanopathies, ciliopathies, cri-du-chat syndrome, pontocerebellar hypoplasia, Aicardi-Goutieres syndrome, patterning defects, pontine tegmental cap dysplasia and other rare disorders.3) Provide key distinguishing imaging features for these malformations.

Conclusions: This review aims at familiarizing radiologists with complex and challenging cases of brainstem maldevelopment from in-utero to post-natal presentation. Early and accurate diagnosis are key to meaningful outcomes and future planning. It is important for neuroradiologists and pediatric radiologists alike to retain a wide range of differential diagnoses thus providing diagnostic confidence when potential future cases are encountered. Purpose or Case Report: The purpose of this exhibit is explore pitfalls in our experience with SPECT-CT Iodine-123 (I-123) MIBG imaging in patients with neuroblastoma. SPECT-CT can more specifically localize areas of uptake over planar imaging and mitigate false-positive results with correlative anatomic information. We will review cases of false-positive MIBG uptake in nonmalignant sites, cases of false-negative MIBG uptake in neuroblastoma/neural crest tumors, and cases of secondary tumors/malignancies occurring in the setting of known neuroblastoma, with variable uptake on MIBG. Purpose or Case Report: With the increased demand and utilization of PET/MRI in oncologic imaging, there is an ever increasing database of non-malignant pathology that has not yet been described on PET/MRI. Furthermore, the pediatric population provides an even more unique breadth of pathology that is often only seen in this age group. It is important to be able to accurately identify these common pathologies so as to not mistake them for malignancy and to prevent unnecessary follow up imaging studies and further invasive diagnostic procedures.For this educational exhibit we review over 200 clinical pediatric 18F-FDG PET/MRs performed at our institution and highlight the most common and most interesting cases of FDG-avid non-malignant pathology.Listed below are some of the cases to be included in the poster:-Benign FDG avid bone tumors including non-ossifying fibromas-Benign causes of FDG avid lymphadenopathy including cat scratch disease-Benign causes of FDG avid lung lesions including aspiration pneumonia-Benign causes of gastro-intestinal FDG uptake including pseudomembranous colitis-Benign FDG avid infectious pathologies including a liver abscess Purpose or Case Report: Post-Transplant Lymphoproliferative Disease (PTLD) is a polyclonal and monoclonal lymphoid proliferation which occurs in 1-20% of solid organ transplant recipients. It is most common in multivisceral organ transplants followed by small bowel transplants, heart and lung transplants and less commonly in liver and kidney transplants. PTLD has a bimodal distribution of occurrence with the largest peak occurring within 1 year after transplantation and a second peak at approximately 4-5 years after transplantation. The Epstein Barr virus (EBV) is associated in 50-70% of cases. EBV seronegativity in the recipient at the time of transplant predicts a 2-4 times increased risk of PTLD especially if they receive a donor organ which is positive for EBV. This may explain the higher incidence in the pediatric population who tend to be seronegative for EBV. The World Health Organization (WHO) identifies four pathologic categories of PTLD: early lesions, polymorphic type, monomorphic type and classic Hodgkins lymphoma. The majority of PTLD cases are caused by Blymphocyte proliferation in a T-cell depleted environment in the setting of immunosuppression. However, there is a subset of cases that are caused by T-cell or natural killer cells as well as cases that occur in the setting of negative EBV. Multiple clones of proliferating B-cells can be seen in a single patient. 2/3 of cases have diffuse expression of CD20 which is an important target for therapy.PTLD may be focal or diffuse and can manifest in a variety of different organ systems or even in the allograft itself. There is a higher percentage of extranodal disease in PTLD as compared to immunocompetent patients with lymphoma. The GI tract and liver are most commonly involved. Isolated lymph node involvement in comparison is less common in patients with PTLD. Central nervous system (CNS) involvement is relatively rare in PTLD.This educational exhibit will provide a pictorial review of PTLD and illustrates cases from one of the busiest transplant centers in North America to highlight the major imaging findings as well as complications seen on imaging of this disease. Extranodal and nodal disease will be demonstrated on multiple modalities as well as complications of this disease including intussusception and biliary obstruction. CNS disease will also be shown. The clinical manifestations, imaging characteristics, prognosis and treatment will be discussed and depicted.

Purpose or Case Report: With research pushing ever onward, it is often difficult to keep pace with the dynamic landscape of pediatric abdominal tumors and their classification systems. However, it is imperative that we, as radiologists, remain vigilant of these changes, as our initial and follow-up imaging assessments often have the potential to drive clinical intervention in widely differing directions.In this educational poster, we will review the most up- Purpose or Case Report: Molecular biology has come to the forefront of modern oncology. Knowledge of specific genetic mutations within tumors drives prognostic information and can guide therapy. Keeping up with new terminology in oncology can be difficult for pediatric radiologists who often host oncology conferences and present regularly at grand rounds. The purpose of this educational exhibit is to describe DICER1 syndrome and discuss its implications for pediatric radiology.

Methods & Materials: Basic information regarding the molecular basis for tumor promotion is presented. The DICER1 gene and its product are described. Neoplastic associations with DICER1 are emphasized. The effects on multiple organ systems are discussed.

The DICER1 gene encodes a protein that controls production of micro RNA molecules (miRNA). miRNA serves a regulatory role in gene expression by binding messenger RNA (mRNA). Messenger RNA is the intermediary between the genetic information encoded by DNA and the proteins and enzymes that eventually produce phenotypes. The DICER1 product thus serves as a policeman of sorts, and common DICER1 mutations result in a dysfunction and loss of regulation which increases the likelihood that a neoplastic process will ensue. Most individuals with a DICER1 mutation do not develop cancer, but the risk is increased. Pleuropulmonary blastoma is a primary concern in DICER1 syndrome. Cystic nephroma is also a common association. Sertoli-Leydig cell tumors and multinodular goiter are seen in patients with DICER1syndrome. Pineoblastoma and pituitary blastoma are seen with specific DICER1 abnormalities. Conclusions: DICER1 syndrome is a cancer predisposition condition that can affect multiple organ systems. Understanding the function of this gene is essential to appreciating its associated disease processes. Purpose or Case Report: PET/MR is a valuable and growing imaging method for the assessment and management of pediatric bone tumors. Although plain radiography remains the first line modality for initial evaluation, cross sectional imaging is often required for further characterization of indeterminate or aggressive appearing lesions. Due to its superior soft tissue contrast resolution compared to CT, MR has become the mainstay in tissue characterization, locoregional staging, and surgical planning of pediatric bone tumors. By adding functional and metabolic information, FDG-PET imaging is useful for "one stop" local tumor and whole-body staging, evaluating response to therapy and surveillance. 18F-FDG PET/MR scans have the benefit of lower radiation and increased patient convenience compared to 18F-FDG PET/CT scans. However, due to the relatively recent development of this technology, many radiologists may be unfamiliar with the technical considerations and interpretation pearls and pitfalls of PET/MR. This educational exhibit reviews the imaging technique, reporting requirements, and imaging characteristics of the most common pediatric bone tumors with 18F-FDG PET/MR.

We conducted a comprehensive literature search on 18F-FDG PET/MR of pediatric bone tumors and have included the most current evidence-based information for review. We also present our institutional approach and experience with performing 18F-FDG PET/MR scans of pediatric bone tumors. Results: We describe the imaging technique and reporting criteria for conducting 18F-FDG PET/MR scans of bone tumors in children and young adults. We review the epidemiology, pathology, 18F-FDG PET/MR imaging characteristics, and treatment monitoring approaches for the most common pediatric bone tumors, including osteosarcoma, Ewing sarcoma, primary bone lymphoma, bone and bone marrow metastases, and Langerhans cell histiocytosis. We also discuss various potential "false positive" bone lesions, and some important similarities and differences between 18F-FDG PET/MR and 18F-FDG PET/CT. Finally, we provide insight into the future directions and developments of this new technology. Purpose or Case Report: Childhood interstitial (diffuse) lung disease (chILD) in infants consists of a rare and heterogeneous group of disorders previously classified with clinical, radiologic, and pathologic features. The purpose of this article is to discuss imaging techniques and provide a pattern-based approach for chILD in infants.

1. Review the current American Thoracic Society (ATS) guidelines for diagnosis and classification of chILD2. Discuss the utility and limitations of imaging modalities including radiographs, CT and MRI for diagnosis and follow up of chILD in infants3. Review a CT pattern-based approach with imaging examples for chILD in infants method was employed to assess collimation: a horizontal line was drawn across the image at the level of the inferior border of the lower hemidiaphragm, an "optimal" vertebral body level was assigned by allowing 1 vertebral body height below the line for patients 5 years and younger (0.5 vertebral body heights for patients older than 5 years), and comparing this to the actual lowest vertebral level included in the film. After analysis of the control group was complete, an in-service was provided at the community hospital in which the proper and expected technique was reviewed with the technologists. Results: In the initial analysis of films in the pre-intervention group, an average of 1.72 extra vertebral levels were included on 49 frontal radiographs analyzed. A single film was excluded due to over-collimation meaning that both costophrenic angles were not included. Of the 39 lateral films obtained in the control period, an average of 1.33 extra vertebral levels were included. Of the lateral films, 2 were excluded due to over-collimation.

Reducing unnecessary x-ray exposure to children through the use of an inservice program is achievable. Purpose or Case Report: The principle of ALARA and its application in the pediatric population serves as the focus of the Image Gently campaign. Tertiary care institutions dedicated to the care of children are well situated to be regional leaders in maximizing the ALARA concept. This study compares the sizespecific dose estimates of computed tomography (CT) studies performed at our institution to those from outside facilities referring patients for emergent care.

Our institution, a major pediatric referral center, is a large academic pediatric hospital that performs approximately as many CT exams of the abdomen and pelvis with contrast for acute abdominal pain as it receives consult requests for CT exams performed at referral centers. The SSDE of 20 consecutive contrast enhanced CT exams of the abdomen and pelvis performed at our institution were calculated and compared to 20 consecutive CT exams of the abdomen and pelvis with contrast submitted from outside referral facilities over the same time period. Size specific dose estimates (SSDE) of CTs of the abdomen and pelvis were calculated using established reference ranges based on patient size as defined by the AAPM 1 . To establish dose savings, a control scan was performed with the same technique as the shielded scan but with the breast shield removed. Each phantom was scanned a third time without a shield but with global mAs reduction enabled to match image quality at the heart established during the shielded scan. Parameters for the third scan included a liver region boost feature to improve image quality in the abdomen. Entrance skin exposure was measured at the anterior and posterior chest, with the anterior measurement representing breast organ dose. Image quality was assessed using standard deviation measurements in the heart and liver regions.

In the smallest phantom (water-equivalent diameter, WED 16.6 cm), the breast shield provided greater breast dose savings (14%) than global mAs reduction (7.6%) at the same image quality. In the larger two phantoms (WED 23.5 and 28.9 cm, respectively), breast dose savings with the breast shield (20.6% and 18.7%) were comparable to those when using global mAs reduction with the liver boost enabled (18.6% and 18.9%), with similar image quality in the heart and liver.

Conclusions: For small patients and for scanners without region-specific boost features, breast shields provide the best dose savings while maintaining abdominal image quality and therefore should be used for all CT CAP protocols. In larger patients with access to scanners with these advanced dose modulation methods, global mAs reduction can be used to achieve similar results to breast shields. Purpose or Case Report: This study evaluates the value of radiologist-driven imaging education in a pediatric residency program. The primary goals of this educational program were to provide pediatric residents with resources such as the American College of Radiology Appropriateness Criteria, support optimal resource utilization and patient care, increase resident understanding of radiation risk, and determine the value of integrating radiologists into pediatric education.

A needs assessment was performed in which the chief residents of a large pediatrics program were surveyed. The consensus of chief residents was that a 4-part lecture series delivered by a pediatric radiologist would be beneficial to pediatric residents. Topics included general radiation risk as well as basic imaging topics in the chest, abdomen, neurologic system, extremities, and vasculature. Each lecture integrated appropriate ordering, ALARA/Image Gently, and basic image interpretation. Pediatric residents were given a 10-item quiz before and after the lecture series assessing their knowledge regarding the best test to order in clinical scenarios. Residents were also asked, using a Likert scale, to rate their understanding of radiation risk, the ACR Appropriateness Criteria, and other topics of interest before and after each lecture. There was no correlation between PRF and RV strain but RV end diastolic volume (RV EDVi) and stroke volume(RVSVi). Duration time had obvious negative correlation with LVEF but positive correlation with RV EDVi and RV end systolic volume (RV ESVi). Each segment radial and circumferential strain had obvious correlation. Conclusions: Pulmonary regurgitation fraction will cause right ventricular volume overload and ventricular deformation. RV strain was not found to have correlation with right ventricular volume overload and ventricular deformation in this small group study, but more cases need to be enrolled and control group should be include to compare which will be continued to study. Purpose or Case Report: Diseases of the coronary arteries are rare in children, but can present as unexpected catastrophic events without preceding symptoms. It is essential to identify patients at risk to prevent potentially serious cardiovascular events. Coronary CT angiography (CCTA) has been widely studied in adults, but its clinical applications are not well established in children.

Retrospective chart and image review of coronary artery computed tomographic angiography from 2016 -2018 was conducted to assess 1) image quality, 2) unexpected events, and 3) radiation dose. Results: A total of 32 cases of CCTA were performed in our hospital from January 2016 to May 2018 (from 0.5 to 30 years of age; median 12). Indications for CCTA include 1) congenital coronary anomalies (n = 12), 2) after arterial switch operation (n = 10), 3) other congenital heart disease (n = 6), and 4) miscellaneous (n = 4). All patients had a diagnostic echocardiogram. Although a few, initial studies were performed on a 64 slice scanner, most of the studies were performed using a 2 x 128 slice dual source scanner. High-pitch, prospectively gated, adaptive, prospectively gated, and dose modulated, retrospectively gated techniques were utilized depending upon the patients' heart rhythms. Exposure parameters, kilovolt peak (kVp) and milliampere-seconds (mAs) were adjusted according to patient size and body habitus, as well as due to the presence of metallic chest implants. Three had suboptimal CT images due to an ectopic beat (1), hiccups (1), and a limited injection rate (1). Twenty nine cases (91%) showed expected diagnostic quality: 97% for coronary ostia and 91% for distal segments, superior to echocardiogram. Average/median radiation dose was 2.83/1.97 mSv, which was improved in 2018 (1.53/1.62 mSv) after protocol revision. Conclusions: Coronary CTA provides more reliable image quality of the coronary ostia and distal segments than echocardiogram and delineates a distinct spatial relationship with extravascular structures that angiography cannot. The radiation dosage has been significantly reduced with proactive heart rate control by pre-and intra-procedural use of betablockers, as well as implementation of aggressive protocol dose reduction techniques. CCTA is a reliable and safe diagnostic modality in analyzing coronary anatomy in children. Purpose or Case Report: Duchenne muscular dystrophy (DMD) is a myopathy with a natural history of progressive cardiomyopathy and vasogenic edema, fatty infiltration, and myocardial fibrosis. We recently identified a group of patients with DMD who presented with acute onset of chest pain (ACP), troponin leak (TL), and new late gadolinium enhancement (LGE), similar in presentation to acute viral myocarditis (AM). It is unclear if these patients have suffered an episode of AM or if this is a different disease process. Given the presumed acute myocardial necrosis in both processes, we hypothesize that native T1 and T2 values would be elevated in AM and DMD patients with ACP compared to asymptomatic DMD patients.

This was a retrospective study of three groups: Group 1: DMD, LV ejection fraction (LVEF) > 55%, and no LGE; Group 2: DMD presenting with ACP, TL, and new LGE; Group 3: clinical diagnosis of AM. All patients from January 2016 to June 2018 with native T1 and T2 mapping performed during a clinically indicated cardiac MRI study were included. T1 and T2 mapping was performed at a single midventricular short axis slice. The slice was divided into a septal, anterior, and inferior region. All images were post-processed by a single observer. Results: LVEF was lower in the Group 3 (p <0.05) compared to Group 1, with no difference between Group 2 and 3 (p = 0.99).

Native T1 values were elevated in Group 2 (p < 0.01) and Group 3 (p = 0.02) compared to Group 1 with no difference between Group 2 and 3 (p = 0.41). However when comparing T2 values for Group 1 and Group 2, this did not reach statistical significance (p=0.72) despite their similar age. T2 values for Group 3 were statistically higher for all segments when compared to Group 1 (p < 0.01) and approached statistical significance when compared to Group 2 (p = 0.07). Regional differences were identified in the native T1 maps, particularly in the inferior segment. Conclusions: Patients with DMD who present with ACP, TL, and new LGE have increased T1 values compared to asymptomatic DMD patients, and similar to patients with AM. Increased signal is particularly noted in the inferior and anterior segments. However their T2 values are similar to the asymptomatic DMD patients and lower than AM patients. Although the presentation and clinical course of the DMD patients with ACP and TL has many similarities to AM, the T2 values suggest a lack of significant myocardial edema. This may represent a unique and unusual disease process in the natural history of DMD. Conclusions: Utilizing one of the largest reported series, this advanced imaging series of omphalocele demonstrates the value of organ-specific characterization to prognostication. While identifying liver position has long been recognized as valuable in predicting outcome, our analysis found predictive value for spleen and stomach position as well. These markers should be examined in a larger patient cohort with the goal of creating a validated set of prognostic imaging biomarkers.

Fetal MRI findings in congenital high airway obstruction syndrome: comparison with the normal fetus normal lung (NLV), defined as the difference between total lung volume (TLV) and the volume of the lesion. NLV was normalized against published lung volumes by gestational age (O/E NLV). Imaging parameters were correlated with respiratory outcomes at birth including respiratory symptoms at birth, need for oxygen supplementation, mechanical ventilation, and delay in feeding. Purpose or Case Report: Fetal CT has almost the same utility as a postnatal skeletal survey. Despite this benefit, the associated radiation exposure is disadvantageous and radiation dose reduction is mandatory. It is however impossible to measure the actual radiation dose to the fetus directly. Several previous reports have described the CT dose index (CTDI) volume and dose length product (DLP) as representing an imagined fetal dose. The actual fetal radiation dose needs to be confirmed using a phantom that practically corresponds to a pregnant woman.

We created a custom-made phantom that corresponds to a pregnant woman, using acrylic resin (Polymethyl methacrylate: PMMA; 340 (Width) x 260 (Height) x 300 (Length)). It could contained three artificial fetal skeletons with different densities of calcium fluoride (300 HU: 1.90 mol/L, 500 HU: 3.63 mol/L, 700 HU: 5.37 mol/L). Dose measurements were obtained using four CT scanners (GE, TOSHIBA, SIEMENS, PHILIPS) at three different institutions with the same scan parameters (CTDIvol: 3.0 mGy; tube voltage: 80, 100, 120 kV) at four measurement points (distance from skin surface: 4 cm, 7 cm, 10 cm, and center). We compared the radiation dose for each tube voltage, at the different measurement points. The differences in dose between CTDIvol and measured data were assessed. Results: The mean measured dose was 2.3 mGy at the center, and 3.73 mGy, 4 cm from the skin surface, across all vendors and tube voltages, corresponding to doses between 76% and 124% of the displayed CTDIvol (3 mGy). The maximum deviation was 1.58 fold (at 120 kV, distance from skin surface: 4 cm). Doses in the peripheral part of the pelvis showed 1.5-1.8 fold greater exposure in comparison with the central position. Doses increased in proportion to tube voltage settings and mean doses at 120 kV were 1.03-1.18 fold greater than at 80 kV. There were discrepancies in the measured dose among four CT scanners that ranged from 61 to 120% (around the mean of 120 kV). Conclusions: Fetuses undergoing CT for suspected skeletal dysplasia may be exposed to approximately 0.8-1.2 times the displayed CTDIvol on the CT console. We suggest that the CTDIvol roughly represents fetal dose. There is however heterogeneity and the wide range of fetal doses depends on the position of the fetus, the selection of tube voltage, and CT units.

As pediatric radiologists, we should be aware of these characteristics so that we can prevent excess radiation.

Purpose or Case Report: MRS is the best prognostic indicator for hypoxic-ischemic encephalopathy (HIE) but is difficult to obtain early after injury. We investigated whether cerebral blood flow measures of resistive indices (RI) and time average maximum velocities (TAMx) shortly after birth would relate to later degree of neural injury by MRI in hypothermic HIE newborns. We predicted that abnormally high/low blood flow would be associated with poor outcomes.

We retrospectively investigated 81 infants born between 2012 and 2018, ≥34 weeks gestation, treated with hypothermia, who received a transcranial Doppler ultrasound within 24 hours after birth, and MRI at 3-10 days. Cerebral blood flow measures (RI, TAMx) in anterior cerebral (ACA), middle cerebral (MCA) and basilar (BA) arteries were correlated with MRS ratios of neuronal health (Nacetylaspartate, NAA) in the basal ganglia (BG) and frontal white matter (WM). As both high and low RI and TAMx are abnormal, we divided our data into quartiles (Q) to find linear correlations between blood flow and NAA. Results: As resistance in MCA increased above normal (Q3), NAA ratios in WM decreased (r 2 = -0.574, p=0.02), reflecting a decrease in neuronal integrity. Also, as blood flow velocity in BA (Q4) increased above normal, NAA ratios decreased in BG (r 2 = -0.550, p =0.012). Conversely, as velocity in the BA approached normal (Q2), NAA ratios increased in WM (r 2 = +0.618, p=0.011) indicating greater preservation of axons.

In this largest reported sample of cerebral blood flow in hypothermic HIE neonates, increased resistive index in the MCA and cerebral blood flow velocity in BA in the first 24h after HIE birth are associated with more injury, and worse NAA ratios. Transcranial doppler US can be performed at bedside shortly after birth and might prove useful for earlier prognosis in neonates with HIE.

scores, highest percentage of necrotizing pancreatitis, and highest US sensitivity (83%). Conclusions: Most cases of AP in children being treated for ALL can be diagnosed with clinical history and labs. When imaging is used, US is much less sensitive in detecting AP than CT, except in the most severe cases (CTAE Grade 4). Imaging to diagnose AP in this patient population should be limited to clinically equivocal cases.

Feasibility of a scoring system to predict reducibility and surgical outcomes of ileocolic intussusceptions in children. Purpose or Case Report: Determine if a simple scoring system can accurately predict the reducibility of ileocolic intussusceptions with therapeutic enema, and eventual surgical outcomes in those patients failing reduction.

An IRB approved, retrospective study was performed. In previous, unpublished research, the authors identified several imaging findings associated with a significantly decreased success rate of therapeutic enema, including presence of bowel obstruction on plain film, trapped fluid and poor blood flow on ultrasound, and location of the leading edge of the intussusception at or beyond the splenic flexure. A scoring system was devised assigning 1 point to each of these findings. This scoring system was then retrospectively applied to the cases of ileocolic intussusception at our institution between September 2012 -August 2017 in which plain films and ultrasounds were both available. The cases were then stratified according to their scores, and correlated with enema results and surgical outcomes. Complicated surgical cases were defined as those that required more extensive surgery than a standard laparoscopic reduction, such as conversion to open laparotomy. Results: 191 cases were reviewed. Following point assignment, the following categories were generated: Category 0 (114 cases), Category 1 (35 cases), Category 2 (32 cases), Category 3 (8 cases), and Category 4 (2 cases). Category 0 had the highest enema success rate (86%) and lowest rates of complicated surgery (3%) and bowel resection (3%). Lower rates of enema success, and higher rates of complicated surgery and bowel resection, were seen in higher categories. In grouped analysis, there was a statistically significant decrease in enema success rate between Categories 0-1 and Categories 2-4 (83% vs 19%, p < 0.0001), and increase in complicated surgeries (4% vs 40%, p < 0.0001) and bowel resections (3% vs 29%, p < 0.0001). Conclusions: A simple scoring system based on imaging findings obtained in most cases of ileocolic intussusception can predict reducibility by therapeutic enema, as well as surgical outcomes in cases of failed reductions. This information can be useful to radiologists, surgeons, and emergency physicians when determining patient management, as some patients may be better served by proceeding to surgery rather than enema. Further research to validate the scoring system in a prospective fashion is planned. Purpose or Case Report: Determine if the length of time between diagnosis of an intussusception and therapeutic enema, or reported length of symptoms, is associated with lower enema success rates or higher rates of complicated surgery and bowel resection.

An IRB approved, retrospective study was performed. The radiology information system at our institution was queried for the keyword "intussusception" in fluoroscopic reports from September 2012 -August 2017. The length of time ("time to catheter") between diagnosis of intussusception, defined as the end time stamp of the ultrasound or CT on which it was first identified, and the start of therapeutic enema, defined as the first time stamp on images from the enema, was then calculated. In addition, the medical record was reviewed for overall length of symptoms (LOS) as documented in the emergency department note. Correlation with enema results, and surgical outcomes in patients failing reduction, was then performed. Complicated surgical cases were defined as those that required more extensive surgery than a standard laparoscopic reduction, such as conversion to open laparotomy. Results: 209 cases were reviewed. There were 148 successful enemas and 61 failures. No significant difference was found in mean overall LOS between the 2 groups (1.44 days vs 1.58 days, p = 0.44) or in mean "time to catheter" (1 hour 53 minutes vs 1 hour 54 minutes, p = 0.3). Using grouped analysis, in patients with "time to catheter" of 0 -4 hours versus those > 4 hours, no significant difference was identified in enema success rates (71% vs 65%, p = 0.6), rates of complicated surgery (11% vs 12%, p = 0.9), or rates of bowel resection (8% vs 12%, p = 0.56). Similarly, there was no significant difference among groups with the shortest and longest LOS. In patients with LOS 0 -1 days versus those > 2 days, enema success rates (73% vs 68%, p = 0.5), complicated surgery rates (9% vs 15%, p = 0.25), and bowel resection rates (6% vs 13%, p = 0.14) were not significantly different. Conclusions: Air enemas for intussusception performed within 4 hours of diagnosis did not show a difference in rates of enema success, complicated surgery, or bowel resection compared to those performed greater than 4 hours after diagnosis. Additionally, overall length of symptoms does not show a difference in rates. Putty® (Ja-Ru), polyvinyl acetate in water (Elmer's® glue), support material in water (Support 705®, Connex 500) and Metamucil® in water were tied-off into nitrile gloves separately. These were placed individually into the central lumen of the synthetic bowel in a tub of water and a sonographer with 17 years experience scanned the bowel using a Philips EpiQ-7G to evaluate the sonographic characteristics. Microbubbles (Lumason®, Bracco Diagnostics Inc) were then injected into the false lumen and rescanned. A numerical grading system was developed ranging from 4 to 12 to determine the optimal material. The criteria assessed on US included: visualization and imaging quality of the bowel wall(s), ± sonographic artifact and presence of echogenicity on B-mode and signal intensity on contrast mode. The luminal material with the highest score was deemed the optimal material. Results: Materials ranged from a score of 6 to 12. The Play-Doh®, support material and Elmer's® glue each had the lowest scores of 6 due to poor image quality, extensive artifact and poor visualization of the bowel walls and Lumason® on contrast mode. The Kinetic Sand™ had a score of 7 and was better at visualizing the lateral bowel wall. The Metamucil® and Flarp Putty® had scores of 8 and 9 respectively, with complete visualization of the bowel wall and little artifact but both materials had increased echogenicity on B-mode. Play-Doh® in water and Kinetic Sand™ in water had the highest scores, 12 and 11 respectively. These materials demonstrated no artifact from luminal content with optimal image quality of the bowel wall. Yet, the Play-Doh® in water scored the highest due to best Lumason® visualization in the bowel wall.

Conclusions: The preliminary data shows Play-Doh® dissolved in water results in optimal sonographic characteristics for use in a 3D US bowel phantom. Such a phantom model will serve as a valuable training tool for grayscale and contrast enhanced bowel US evaluation. Purpose or Case Report: To assess the incidence, scope, clinical findings and imaging characteristics of GU pathology due to underlying KTS in pediatric patients with the goal of improved diagnosis and outcomes for children with this potentially fatal disorder.

Using a retrospective data analysis design, the charts and imaging studies of 58 patients with confirmed diagnosis of KTS within a single tertiary care pediatric health care system were studied for insight into the degree of GU involvement in pediatric patients.

Results: 17% of patients (10) were noted to have GU findings, with 5% (3) having extensive GU involvement (defined as greater than 1 finding). The overall scope of GU involvement was diverse and varied with more than 9 different findings identified spanning 6 unique organs. Kidney, bladder and scrotum had the most common findings. Hematuria was the most common presenting symptom in 30% (3/10). Findings included those discussed in the literature along with previously unreported lymphatic findings. The importance of early recognition of GU involvement due to KTS was confirmed as the average age of GU complication was 7.6 years, less than 3 years after the average age of KTS diagnosis in this patient population.

Conclusions: Significant GU complications due to KTS can occur in the pediatric population with pathologies manifesting at an early age and with delayed recognition. Early clinical and imaging characterization of these conditions is important for expectant management and early intervention strategies, thereby playing a central role in improving patient education, patient outcomes and reducing costs, morbidity and mortality. Purpose or Case Report: Accurate placement of lines and tubes is essential in pre-hospital emergent care of pediatric patients, to ensure that lifesaving measuring can be initiated in an effective and timely manner. Malpositioned pediatric tubes can lead to failed resuscitation efforts. The purpose of this study is to determine the prevalence of misplaced endotracheal tubes (ETT), nasogastric tubes (NGT) and intraosseous trocars (IOT) in the pediatric acute care setting using post-mortem imaging.

The post-mortem imaging for 87 patients under the purview of the Medical Examiner Office performed between June 2007-May 2017 was retrospectively reviewed by a Pediatric Radiologist as part of a quality review in conjunction with the Nova Scotia Medical Examiner Service and Emergency Health Services (EHS). Of these 87 patients, 27 patients were excluded as no tubes were placed. For the remaining patients (N=61), ETT, NGT and IOT placement was assessed using a combination of radiographs and multiplanar CT, and characterized as "satisfactory position" or "malpositioned". For the malpostioned devices, the exact location was also documented. Results: Of the 45 patients with an ETT, the tube was in satisfactory position in 9 (20%) patients and malpositioned in 36 (80%) patients. Of the malpositioned tubes, 5 were at the carina, 20 in the right main bronchus, 9 in the esophagus and 5 in other locations (at T2, T4, nasopharynx, etc.). The NGT was in satisfactory position in 13 (76%) patients and malpositioned in 4 (24%) patients, with tubes in the pharynx (two cases) and distal esophagus (2 cases). Of the 58 IOTs inserted, 15 (26%) were in satisfactory position and 43 (74%) were malpositioned, including 10 that had been completely removed but with CT evidence of prior incorrect positioning, 8 were through and through bone and 6 in the superficial soft tissues. Conclusions: Malpositioning of the ETT, NGT and IOTs in this study is likely multifactorial. Tube placement is less common in the pediatric emergency setting compared to adult medicine, possibly resulting in decreased experience and comfort level among pediatric EHS teams. There may be a role for further specialized training, with more frequent recertification. Use of a laryngeal airway mask could also be considered as an alternative to ETT in pediatric patients with difficult airways. The results of study will help to inform future training of pediatric EHS teams at our institution, with ongoing review of post-mortem imaging as a means of quality assurance. 

Following institutional review board approval, through semi-structured interviews, researchers explored perspectives of 30 PEDP's on how they seek and obtain imaging consultation and the potential impact of ICDSS on practice. Open-ended questions were designed to elicit a range of responses rather than quantifiable data. The team employed a directed content analysis approach and codes were developed and explicitly defined through team evaluation.

Coding and analysis was facilitated by NVivo 12 Software. Two research assistants independently coded all transcripts and discrepancies were addressed through consensus (mean k = 0.99). Codes were arranged into thematic categories to inform an explanatory model that illustrates the PEDP's priorities and needs related to the imaging decision process. Results: Data analysis revealed three thematic categories: (1) Common Influences on the Decision Process (these include patient factors, PEDP's experience, the specialist perspective, and logistics), (2) Radiology Consultation Experience, (3) PEDP Perspectives on ICDSS which includes perspectives on how implementation of an ICDS system at our institution may impact practice. PEDP's described consultation as a valuable component of ICDS but reported difficulty in coordinating an imaging strategy with radiologists and other physician consultants. PEDP's described the exchange of ideas as especially worthwhile for clinical scenarios that do not fit neatly into ICDS pathways. Significant barriers to radiologist consultation included: time, access to radiology attendings, limited confidence in trainees, and not wanting to disrupt radiologist workflow. PEDP's expressed a belief that ICDSS could potentially improve their workflow and augment effective interaction with the radiology department. Conclusions: PEDP's predict that radiologist consultation will continue to be an essential component of ICDS following the implementation of ICDSS. These results present an opportunity for radiologists to improve in-person radiologist consultation in clinical and didactic settings. Additionally, these findings will inform the development and implementation of an ICDSS system that meets the needs of the PEDP's and ultimately improves patient outcomes. Purpose or Case Report: Peer review is an essential component of quality improvement in radiology departments. Over the last years, several papers described a shift towards peer learning, emphasizing that feedback in the form of comments and peer learning conferences creates a more productive environment for reviewing and understanding perceptual or interpretive errors of peer radiologists, with improved diagnostic performance as the objective.At our institution, we instituted a three-month scoreless peer review pilot period.

Radiologists' impressions regarding the current commentenhanced numerical system and the pilot scoreless, commentsonly period were collected through an online survey.

The survey containing 17 questions was distributed twice to the radiologists in our group to evaluate and compare their perceptions regarding the numerical scoring and the scoreless peer review systems. A Likert-style rating scale of 1-5 was used for the survey, ranging from strongly disagree (1) to strongly agree (5). The first survey, administered at the inception of our pilot period, was applied towards the numerical scoring system. The second survey, administered at the end of the pilot period, targeted the scoreless peer review system. Responses from both surveys were compared. Results: 16 out of 20 radiologists in our group responded to both surveys. Overall satisfaction was slightly higher (weighted average of 3.64 versus 3.5) for the scoreless peer review compared to the numerical scoring system. Our colleagues felt more likely to underreport while reviewing a case with the numerical system as compared to the scoreless one. Reported negative emotional impact from discordances was slightly higher with the scoring system. Radiologists felt more inclined to provide a comment for educational purposes when agreeing with the interpretation while using the scoreless system. Comments were felt to be important in both systems. Conclusions: A scoreless comments-only peer review system shows promise in providing more educational opportunities than a comment-enhanced numerical system, with less negative emotional impact on participant radiologists. This benefit may be even greater in practices in which radiologists are prone to underscore their colleagues. Purpose or Case Report: Two major insurance companies have recently enacted policies requiring outpatient advanced imaging at free-standing imaging facilities for adults. Examining the potential clinical and economic implications of non-subspecialized interpretations in children is important. We evaluated the rate of major discrepancies and changes in management arising from second interpretations by a dedicated pediatric facility.

A retrospective and prospective analysis of CTs and MRs presenting for over-read by a pediatric radiologist was performed. For the prospective review, a standardized dictation template was implemented to track disagreements. A board-certified radiologist identified cases with major discrepancies as defined by those that were likely to change surgical or medical management. A chart review was performed to identify changes in clinical course and outcomes based on the second interpretation. Results: 1397 patients were referred to our pediatric hospital for second interpretation following an outside facility interpretation. 3.6% of all patients (18/574 patients from retrospective review and 32/823 patients from prospective analysis) had major discrepancies between the outside report and the pediatric subspecialty interpretation. Of all major discrepancies, 31 were body cases (chest, abdomen, or pelvis), 13 were neuroradiology cases (head or neck), and 6 were musculoskeletal cases (spine or joint). Second interpretations changed surgical management for 14 patients, with 4 patients receiving a necessary surgery and 10 patients avoiding an unnecessary surgery. Medical management changed for 36 patients. Eleven of the cases with major discrepancies involved an incorrect diagnosis related to appendicitis. Examples of other serious missed findings included secondary reads that lead to diagnoses of meningoencephalitis, Crohn's disease, herpes encephalitis, intracranial hemorrhage, pulmonary embolism, and ovarian teratoma.

Conclusions: Pediatric subspecialty interpretations altered the surgical or medical management in 3.6% of referrals, which compares with a 0.4% rate of level 3 and 4 errors from the ACR RADPEER database. This supports the case for performing and interpreting these studies at a facility with dedicated pediatric radiologists. Since changes in reimbursement related to hospital-based cross-sectional imaging are driven by costs, the potential long-term adverse economic impact of nonsubspecialized interpretations suggested by our study needs to be carefully considered. limitations, 3=minimal limitations, 4=no limitations). To assess cost effectiveness, success rate of avoiding sedation after simulation was calculated along with average differential cost to payer for a sedated MRI vs. a non-sedated MRI after MRI simulator training. Results: 592 patients (mean age: 7.9 ± 3.0 years) were enrolled in the MRI simulator program over 4 years with the majority scheduled for a brain MRI (71%). The mean image quality score for sedated brain MRIs was 3.50 ± 0.51, and the mean quality score for non-sedated brain MRIs after successfully completing the MRI simulator program was 3.37 ± 0.49. All examined sequences demonstrated minimal to no limitations for clinical use. No sequence was found to have moderate or significant limitations for clinical use in either group. The success rate of children avoiding sedation after MRI simulation was 87%. The average differential cost to payer was $963 higher for a sedated MRI than an MRI that avoided sedation after MRI simulator training. Factoring in the success rate, the total payer cost savings over 4 years was estimated to be $495,984 for our simulator program.

Conclusions: In addition to reducing sedation rates, our study validates the clinical image quality of non-sedated pediatric MRIs obtained after successful simulator training and demonstrates the cost effectiveness of a simulator program over 4 years. The clinical quality and large payer cost savings should be considered when evaluating future reimbursements and investments for MRI simulator programs in pediatric patients. (6); 2 were hearing aid batteries (1); 20 were a single penny (1), 3 were 2 pennies stacked on each other, 4 were a nickel (1), 1 was a dime (1), 2 were a single quarter (1), and 2 were a quarter with at least one other coin. Remaining cases were identified only as "coin", were foreign coins, or other round objects.12 clinicians ranging from PGY-2 to 30 years of experience participated in the study: 5 radiology residents, 2 radiology fellows, 2 radiology attendings, 1 ED fellow, 1 ED attending, and 1 pediatric attending. The cases were presented in Synapse PACS. The participants indicated the likelihood the object was a button battery using a Likert scale: 5 (definitely), 4 (probably), 3 (equal likelihood), 2 (probably NOT), 1 (definitely NOT); to decide how urgently the object needed to be removed (4=within 2 hrs regardless of NPO status, 3=within 24 hrs, 2=electively, 1=does not need to be removed); and were asked how comfortable they were identifying button batteries on radiographs before and after viewing the cases and after reviewing the answers. Purpose or Case Report: 3D printing technology presents a unique opportunity for the creation of custom phantoms for training and simulation for pediatric interventional procedures that are complex and/or uncommonly performed. The purpose of this study was to describe the elements of designing a 3D phantom for simulation of pediatric abdominal intra -vascular procedures.

In order to create a phantom design considerations such as the consistency of the "soft tissues", vascular system, creation of flow simulation and visibility of the vascular components using fluoroscopy, needed to be considered. Results: Silicone was the chosen material poured into 3D printed molds with break-away and dissolvable internal cavities to create the anatomic vein and atrial system. The phantom design was also divided into three main regions of interest: chest, neck, and thigh. Rubber tube connections between the regions and pumps allowed for circulatory flow.The design process of the phantom consisted of incorporating several 15year-old patient CTA exams to create the atrial and venous pathways. The pathways were then aligned with regions (chest, neck, thigh) segmented from CT exams in MaterialiseMimics and3-Matic. Connection ports between regions were added. Access points covered with a replaceable "skin patch" were added to the neck and thigh regions for reusability. The mold and breakaway cavities were 3D printed on a StratasysFortus450mcin ABS plastic. Smooth-On Ecoflex30was selected for molding to simulate the body and DragonSkin10for the skin patch based off the ShoreValue.Phantom feasibilityfor fluoroscopic visibility, contrast visibility, angiography, venography and device placement and removal was assessed and was successful in this prototype. Conclusions: Phantom creation for pediatric interventional radiology simulation and training is an exciting prospect in pediatric IR for complex and/uncommonly performed procedures both for attending staff skill maintenance and initial training for fellows and residents. Going forward subsections of this phantom will be removable for repeat procedures such as stent placement. Purpose or Case Report: Aneurysmal bone cysts (ABCs) are expansile lytic lesions and can cause significant disability and pain most commonly seen in patients 10-30 years old. This study evaluates the technical feasibility, safety, and efficacy of percutaneous management of ABCs in pediatric and adolescent patients.

In this Institutional Review Board approved retrospective study, 11 consecutive ABCs treated with doxycycline sclerotherapy and/or percutaneous ablation from December 2015 to October 2018 were evaluated. Demographics, clinical presentation, procedural details, and safety/efficacy outcomes were assessed. Results: A total of 11 ABCs in 11 patients (average age = 13.4 yrs; range: 8-17) were treated with 22 total procedures (range: 1-4 treatments/patient). 7/11 (63.6%) ABCs presented with recurrence following surgical resection. ABC location included spine 3/11 (27.3%), pelvis 4/11 (36.4%), lower extremity 2/11 (18.1%), humerus 1/11 (9.1%), and rib 1/11 (9.1%). Doxycycline sclerotherapy alone was used in 10 procedures in 4 (36.4%) patients, cryoablation alone was used in 7 procedures in 4 (36.4%) patients, and a combination of percutaneous ablation and doxycycline was used in 5 procedures in 3 patients (27.3%). All 22 (100%) procedures were technically successful. Clinically, 10/11 (91%) showed clinical improvement including cessation of pain and improved range of motion. Mean followup time= 9 months (range: 1-20 months). On imaging, 9/10 (90%) ABCs decreased in size with increasing sclerosis on follow-up imaging. 1/10 (10%) ABC did not respond to percutaneous treatment and required repeat surgical resection. One patient is awaiting imaging follow-up. 1/11 (9.1%) patient experienced a major complication of left lower-extremity paralysis, bowel incontinence, and fecal incontinence following treatment of an L4 ABC that resolved. 1/11 (9.1%) patient experienced a minor complication of transient numbness of the lateral thigh. Conclusions: Percutaneous treatment of ABCs is a safe and efficacious treatment option in the pediatric and adolescent population.

Posteriorly Tunneled Central Lines in Children: Does it decrease rates of dislodgement?

Michael Acord, MD 1 , mra0305@gmail.com; Carl A. Termine 2 , Anne Marie Cahill 1 , Fernando Escobar, MD 1 ; 1 Radiology, Children's Hospital of Philadelphia, Philadelphia, PA, 2 Monmouth Medical Center, Long Branch, NJ

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: To assess the rate of dislodgement of posteriorly tunneled central lines, over-the-shoulder, in children.

This was a single center, IRB-approved retrospective study of all children with a posteriorly tunneled central line placed in pediatric interventional radiology over a 12-year period. The following parameters were reported;

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Aneurysmal bone cysts are benign expansile osseous lesions which may behave locally aggressively, leading to growth plate destruction, angular deformity, and pathologic fracture. Current treatments such as curettage and bone graft reconstruction, sclerosis, and intraarterial embolization have high recurrence rates and morbidity. Percutaneous cryoablation is a promising new treatment modality. Three patients with ABCs underwent cryoablation at our institution and follow-up MRIs over the span of a year were obtained. This is the first study of its kind to elucidate the imaging findings post-cryoablation of ABCs, which may help the radiologist interpret response to treatment.

In this HIPAA compliant, IRBapproved study, three children with ABCs were treated using IcePearl percutaneous cryoablation needles (Galil Medical).

Patients included an 11-year girl with a clavicular lesion, a 16 year-old girl with a femoral lesion, and a 14-year old boy with a femoral lesion. Clinical and imaging follow up with MRI occurred at 3, 6, and 12 months. Results: Common imaging findings after cryoablation include resolution of fluid-fluid levels, decreased size of a T2hyperintense core, appearance of a sclerotic rim of bone, and a cryoablation burn zone. All patients reported markedly decreased pain and return to normal activity by 6 months post treatment, including one patient able to play a full hockey game at 3 months. No major complications were reported. No clinical or imaging indicators were present to suggest recurrence. Conclusions: ABCs are a difficult entity to treat with high recurrence rates and morbidity with current treatment options. After percutaneous cryoablation, ABCs may show common imaging findings which may help radiologists monitor treatment efficacy. All three patients treated at our institution showed excellent clinical results. Percutaneous cryoablation is a promising treatment modality for ABCs which in our series confers exceptional benefit to patient symptomatology without serious morbidity. Purpose or Case Report: Acoustic radiation force imaging (ARFI) is a type of quantitative sonoelastography (SE) that has been used to determine tissue stiffness in many disorders including liver fibrosis, breast cancer, thyroid nodules and more recently, cutaneous scleroderma. However the use of this technology for the assessment of skin lesions in the paediatric population has not yet been investigated. The purpose of this study is to test the feasibility of using ARFI SE to quantify the stiffness of morphea or localized scleroderma (LS) and hypertrophic burn scars (HTS), two skin conditions in which there is excessive deposition of collagenous and noncollagenous extracellular matrix causing fibrosis.

We employed conventional ultrasound (US) and ARFI SE to characterize changes in the skin of preassigned LS and HTS lesions and compared them to the normal contralateral sites as controls. The inclusion criteria was clinical diagnosis of the corresponding skin condition by a specialist. For each participant, a "target lesion" was selected and marked by the clinician as well as a section of contralateral matched normal skin for US and ARFI measurements. The Wilcoxon signed-rank test was used to compare scar and control sites. Results: 26 patients from the outpatient Morphea (n = 13) and Burn (n = 13) clinics at our tertiary pediatric hospital were prospectively recruited (age range: 4 to 16; 12 females, 14 males). Lesions varied in location and severity. As expected LS lesions and HTS were significantly stiffer than normal skin (E means of 17.61 for LS and 41.11 kPa for HTS respectively, versus 10.32 and 10.05 kPa for the normal controls).Skin thickness of the LS lesions was 30.7% thinner than the respective healthy skin, but the difference between means was only significant for the dermal layer. HTS were significantly thicker than the control sites (p < 0.05). Mean thickness values for the scar and control sites were 4.06 mm and 1.43 mm respectively. Variable changes in echogenicity and vascularity were seen in both group of lesions.

Conclusions: This study demonstrated the feasibility of using ARFI SE to discriminate between normal skin and fibrosclerotic skin conditions (LS and HTS) by measuring skin stiffness. In addition, the findings show that both these types of lesions are significantly stiffer than normal skin. Future research should focus on establishing reference data and determining the technology's ability to detect scar changes over time and evaluate response to treatment. Infections were established using a combination of positive cultures, elevated inflammatory markers, clinical assessment, and response to antibiotic treatment. MR studies were retrospectively reviewed by 2 radiologists in consensus for the presence of bone marrow edema, bony erosions, joint effusion, extracapsular edema, soft tissue abscess, and sciatic neuritis. Pre-treatment radiography was reviewed for the presence of radiographically visible osseous change. Clinical chart review was performed for clinical history and outcomes. Descriptive data is presented as mean ± SD. The inter-observer agreement was evaluated with weighted-kappa. Kappa scores (k) of 0.41-0.60, 0.61-0.80 and ≥0.80 were regarded to be indicative of interest (ROI) are created.The CT and MRI sequences are fused using the image registration algorithms in MIMICS TM .VSP and creation of patient specific surgical guides:The registered 3D models are imported into Geomagic Freeform (3D systems,SC,USA).The program has a force feedback haptic phantom and tools for cutting, moving, rotating, and creation of offsets, etc. Resection(osteotomy)planes are created on the model after due consideration of the safety margin required for the type of tumor,blood vessels in the line of osteotomy, preservation of musculoskeletal function, and any other clinical requirements of the surgeon.The planes are then used to cut the models, simulating the surgical procedure.Open type or closed guides are created depending on the anatomical location,tumor type,surgical plan and surgeon's preferences.Guides are designed as an offset of the remaining normal bone along the osteotomy plan. Screw holes for fixing the guide to the anatomical region of resection are provided.Bone thickness in the ROI is provided ahead of surgery for appropriate planning and preparedness.3D Printing:Designed guides are 3D printed with autoclavable materials such as DentalSG™resin,ULTEM 9085, and Polyamide or with MED610-Sterrad sterilization.Patient specific surgical guides so produced are used in the sterile field and fitted on the ROI.The VSP resection plan is transferred to the OR and executed precisely. Results: We utilized this procedure for successful resection of Ewing's Sarcoma of the scapula in a 12-year-old, preserving as much of the patient's native scapula as safely feasible.

Conclusions: This process can lead to better safety and quality procedures, eventually leading to better clinical outcomes.

Can zone of provisional calcification (ZPC) imaging predict physeal outcome after fracture? Purpose or Case Report: Physeal injuries can result in premature physeal fusion. MR imaging is useful for mapping these bony physeal bridges. Resection of the bony bridge does not reliably restore normal physeal function. We have observed subtle loss of the normal low signal intensity line of the zone of provisional calcification (ZPC) extending beyond areas of physeal bar formation in some patients with prior growth plate fractures. This loss is a marker of disrupted endochondral ossification and likely reflects a more extensive region of physeal damage that might be used to better predict treatment outcomes. Given the subjectivity of visual ZPC assessment, we sought to develop a quantitative 3D map of the periphyseal area of the distal radius using a high resolution 3D fast/turbo spin echo sequence.

Evaluation of two patients' wrist MRI was performed retrospectively with IRB approval. One patient had a prior Salter-Harris type II fracture of the distal radius with a bony bar while the second had a normal wrist MRI performed for non-traumatic reasons.We developed a program to quantify the MRI signal intensity of the periphyseal region in cross section using a semi-automated approach to provide a more accurate assessment of loss of the ZPC. The normally continuous higher signal band of healthy distal radial physeal cartilage adjacent to the normal low signal intensity ZPC was assessed and compared to the normal distal ulnar periphyseal region. Signal metrics and periphyseal 3D signal maps were obtained. These were evaluated and compared.

Results: Four signal metrics were calculated producing associated signal maps that were color coded: Difference in signal minimums, peak minus minimum ratio, slope of minimums and voxel distance from ZPC to peak. Comparisons were made between the distal radius and ulna and between the 2 subjects with fracture and normal wrist. Signal maps demonstrated visually detectable differences that correlate with loss of the ZPC surrounding the physeal bridge versus normal radius and ulna. Conclusions: Quantifiable changes in the ZPC can be demonstrated with this model. The model will be applied to MRI's in patients with prior growth plate fractures and compared with clinical outcomes to determine its utility in providing a more complete assessment of physeal function. Purpose or Case Report: MRI interpretation of pediatric bone marrow is a challenging task due to the highly variable appearance. The bone marrow composition, especially the cellularity, changes with age, benign and malignant hematologic conditions, medications, among other etiologies. Detection of a marrow replacement process on MRI can be missed even by experienced radiologists. Normal hematopoietic bone marrow from birth to 9 years ranges from 20-40% fat (inversely related to cellularity). In malignant marrow replacement processes such as leukemia, bone marrow is highly cellular and fat percentage is low. Currently, no technique is routinely used in clinical practice to quantify bone marrow fat percentage on MRI.

A pilot study of patients with suspected malignant marrow replacement recruited by convenience sampling. MRI and bone marrow biopsy were performed within 8 days of one another. Bone marrow biopsy fat % was averaged between two blinded pathologists. MRI obtained from the suspected area of symptoms and core bone marrow biopsy obtained (as standard of clinical care) from the iliac crest. The MRI technique used was the Dixon technique with fat and water separation. Fat fraction is calculated based on the ratio of fat and water proton density. MRI was considered predictive of malignancy if fat fraction % was less than 20. Results: Six cases of patients (mean age 7.5y (range 3-14y)) were recruited. Pathologic diagnoses included: leukemia (n=5) or rhabdomyosarcoma (n=1). MRI was performed and analyzed from the: pelvis (n=4), spine (n=1), and elbow (n=1). MRI fat% correlated with pathology, but due to the small sample size was not significant (r=0.6, p=.20). Inter-pathologist correlation was high (r= 0.9, p<0.05). Elbow MRI fat% correlated with pathology (76 v. 67.5%), but was falsely negative in predicting leukemia. Of all other locations, mean fat percentage calculated by MRI versus pathology was similar (average± standard deviation, 3.7± 1.3 v 3.0± 3.0). Conclusions: MRI can help radiologists detect a malignant marrow replacement process with more confidence using the fat fraction technique. Studies are ongoing to compare these cases to control patients. differences in seizure duration and seizure types among the three groups. Conclusions: Patients with concordant distribution of the findings between SWI and EEG had significantly frequent multiple or epileptic seizures. SWI may be helpful to discover the epileptic focus localization in children with acute seizures. Purpose or Case Report: There is little published research on the wide variation of corpus callosum (CC) morphology in children, the assessment of which is made difficult by the complex alteration of its appearance in childhood. The purpose of our study was to assess the morphology of the CC on midsagittal T1-weighted magnetic resonance imaging (MRI) in a large number of children and correlate the findings with demographic and clinical criteria.

We reviewed all mid-sagittal T1weighted brain MRI's performed from July to December 2015 and obtained relevant demographic and clinical information from the accompanying report and laboratory system. The CC morphology was analysed by three radiologists and compared using cross tabulation with the chi-square test and ANOVA. Interobserver correlation was assessedusing Kappa coefficient of conformance. Results: 257 children (mean age 72±60 months) were included, (142 male; 55%). In abnormal MRI's the CC was less likely to have an identifiable isthmus and more likely to be convex, thin and have separation of the fornix insertion (all p<0.01). In young children (< 5 years) the CC was also less likely to have an identifiable isthmus (p=0.01) and was more likely to be convex (p=0.04) but the fornix was more likely to insert normally (p<0.01).

Conclusions: There is a distinct pathological appearance of the CC. The immature appearance of the corpus callosum can mirror this but is distinguished by normal insertion of the fornix and normal quantitative measurements.

A decrease in hematocrit in the preceding 48 hours is not predictive of an ICH on neonatal HUS. Those with ICH on HUS had a significantly lower gestational age and were more likely to be born premature.

Poster #: SCI-056 Purpose or Case Report: The diagnostic approach to fever of unknown origin (FUO) in a neutropenic child is challenging.

Recommendations of the Children's Oncology Group and the ACR provide some guidance, but evidence basis for use of sinus CT in the acutely febrile neutropenic child is poorly established. This presentation assesses baseline paranasal sinus mucosal thickening in children in 2 separate cities and compares them to children with neutropenia undergoing sinus CT in the work-up of FUO.

Data collected from 2 large children's hospitals in major metropolitan cities in the southern US. Hospital A is in a desert climate, Hospital B is in a humid subtropical climate. 18 consecutive sinus CTs were reviewed in neutropenic children undergoing diagnostic evaluation for FUO at Hospital A. Bone marrow transplant patients were excluded. All children were oncology patients undergoing treatment. Control groups used include 18 consecutive patients at hospital A and 18 consecutive patients at hospital B who presented to the ED requiring CT of the face. Control patients with a history of sinusitis, patients with a facial bone fracture, and patients with any oronasal support devices were excluded. No exams of neutropenic fever patients at hospital B were available, hospital B uses nasal endoscopy to assess for sinus disease in this population. Lund-Mackay (L-M) and modified Lund-Mackay scores were applied to all CT scans. Results: The mean age in the neutropenic fever group was 9.86 years. The mean ages of hospital A and B control groups were 8.16 and 8.34 years, respectively. Age differences were not significantly different using a Student t-test. LM scores in the hospital A control group exhibited a mean of 5.50 with a mean modified-LM score of 6.92. Corresponding values in the hospital B control group were 3.11 and 4.00, respectively. In the neutropenic fever group of hospital A, the mean LM and modified-LM scores were 4.17 and 4.59, respectively. Findings viewed as "positive" in the neutropenic fever group resulted in 2 infectious disease and 2 ENT consults. All 4 consults concluded that fever was unlikely to be due to paranasal sinus disease, no intervention recommended. Conclusions: Mucosal sinus thickening is omnipresent in children and can vary based on region. Baseline paranasal sinus mucosal thickening is greater among children living in a desert climate. The description of paranasal sinus mucosal thickening in neutropenic children may be of little value in the approach to fever without source. Purpose or Case Report: Detection of focal cortical dysplasias (FCDs) can be challenging. Our aim was to evaluate quantitative contrast differences between FCDs and the mirrorlocation normal gray matter, and the effect thereof on the ease of detection by radiologists.

Coronal FLAIR and T1WI of 20 pathology proven cases of FCD, including 11 cases where the diagnosis was missed on the initial review, were analyzed. CNR and conspicuity differences between gray matter within FCDs and contralateral mirror image locations were calculated using manually drawn ROIs. Similar calculations were made in 20 age and gender matched controls for lateralized differences in normal gray matter at locations equivalent to those of FCDs. Results: On FLAIR images, median (25 th percentile, 75 th percentile) CNR and conspicuity for FCD was 11.3 (8.5, 20.9) and 10.2 (7.1, 15.4), significantly higher than corresponding CNR of 1.0 (-0.2, 5.5) and conspicuity of 0.8 (-0.2, 2.8) for controls (p<0.0001 for both). Similarly, CNR and conspicuity for FCD was also higher than controls on T1WI (p<0.0001). Subgroup analysis showed that these differences from normal were maintained irrespective of FCD detection or lack thereof at the initial review. Median CNR and conspicuity on FLAIR (p = 0.79 and 0.07 respectively) and T1WI (p = 0.64 and 0.76 respectively) were not significantly different for FCD cases that were detected and those that were missed on the initial review. Conclusions: Focal cortical dysplasia lesions have significantly higher CNR and conspicuity relative to mirror-location gray matter when compared to age and sex matched controls. These differences exist even in cases that were missed at the time of initial review. Accentuation of such differences may serve as a means to improve detection of FCDs. Purpose or Case Report: Establish and evaluate a process for 3DPrinting (3DP) patient specific functional anatomical models of the brain from multiple MRI sequences in patients with epileptic focal lesions.The second objective was to evaluate the clinical utility of the models.

Eight patients with intractable epilepsy and focal lesions in the brain were selected for the study.Data acquisition:T1 and T2 weighted MR images with DTI fiber Tractography were acquired on GE Medical systems, Siemens Prisma and Skyra MRI scanners with a field strength of 3 TESLA.Slice thickness range-1.00-1.6mm and the slice increment was 0.5-1mm.Image data processing:Using the segmentation,region growing and 3D modeling algorithms of MIMICS TM (Materialise, Belgium)3D virtual models of the structural anatomy of the affected hemisphere and the lesion were created. DTI fiber tracts and the vasculature were segmented from Dynasuite TM (InvivoInternationalUSA)and exported as DICOM data.3D models of the vasculature and the DTI FIBERs were created and integrated into the structural model of the brain.3DPrinting:After preprint preparation the with Geomagic Freeform(3Dsystems,SC,USA)the model were printed with Stratasys PolyjetCONNEX 3 Objet 350 multicolor printer.Models were post processed with high-pressure waterjet.Model verification:The 3DP models were scanned with Toshiba Aquilion CT scanner with a slice thickness and slice increment of 0.5mm.The Dicom image data was imported into MIMICS TM and a 3D model created.The model so created was overlayed on the virtual model created from the MRI data. Surface variation of the 3DP model and the virtual model created from MRI data was mapped with 3Matics TM .Models were resliced into 2D contours and overlayed on the MRI image dataset for verification.3DP models of the brain were used for team discussion,resident and fellows and patient education of the specific neuro anatomy.

Results: Comparative analysis of the CT 3DP model and the 3D models created from 2D MRI image data showed a mean difference of -0.36 mm with STD Dev.+/-0.5mm. Conclusions: Precise patient specific functional anatomic brain models can be created from multiple MRI sequences using 3D modeling and Polyjet printing. 3DP patient specific models are a valuable adjunct to neurosurgical planning. Brain models help surgeons to educate the patients, parents, residents and fellows to understand neuroanatomy better.Parents understand the clinical situation and suggested treatment implications better and give better informed consent. Purpose or Case Report: Convolutional neural networks (CNNs) have proven to be valuable in the fields of image processing and computer vision. Our work applies complexvalued CNNs to magnetic resonance imaging (MRI) to reduce scan times. The reduction of scan times has widespread pediatric benefits. A typical scan requires that patients remain still for up to an hour to produce a clear image, which is difficult for children without inducing anesthesia, which carries risks. A need exists for greatly improved MRI scan times without the loss of diagnostic accuracy. This scan time can be reduced by subsampling in k-space. We use CNNs to reconstruct images from these undersampled acquisitions. Our work investigates complex-valued CNNs for image reconstruction in lieu of two-channel real-valued CNNs.

Recent work suggests complex-valued CNNs could be more accurate than real-valued CNNs when dealing with complex-valued data. Typically, complex-valued data is fed into CNNs by using a 2-channel architecture where the channels contain the real and imaginary components of the data. This does not preserve phase information, which is valuable for many MRI applications. Recent work in applying complex-valued CNNs to music transcription and speech prediction tasks demonstrates complex-valued models are highly competitive with their real two-channel counterparts (Trabelsi et al., 2018) . Complex-valued neural networks have been applied to MRI fingerprinting with improvements in accuracy in comparison to real models (Virtue et al., 2017) .We apply complex-valued CNNs to subsampled image reconstruction by modifying components of our current CNN within our deep unrolled architecture to be complex-valued. We perform complex convolution and explore various complexvalued activation functions which keep the pre-activated phase intact, as well as activation functions which are based on the phase component. We evaluate the performance in terms of accuracy of this complex model compared to its real counterpart. Results: We trained two otherwise identical CNNs with realvalued convolution and complex-valued convolution, and approximately 900K trainable parameters each. The validation loss was 0.693 and 0.639, respectively. This shows complexvalued networks have the potential to be much more accurate than their real-valued counterparts. Conclusions: Our work shows potential for reducing MRI scan times by more accurately reconstructing images from subsampled data acquisitions using complex-valued CNNs. Purpose or Case Report: To determine if diffusion weighted imaging (DWI) adds value to PET/MRI in pediatric solid tumors.

After obtaining IRB approval, all pediatric simultaneous PET/MR studies with DWI performed between April 2015 and August 2017 were concurrently interpreted by 2 blinded readers (board certified pediatric and nuclear medicine radiologists) in 2 sessions. In session 1, the 2 readers had access to whole body PET, Dixon, and HASTE images. Images were assessed for quality, number and location of lesions, designation as malignant or not. After 6 weeks, in session 2 the above images were re-reviewed by same readers with DWI added, for the above assessment with the added value of DWI. The results of the two sessions were compared by an independent reviewer. Results: PET/MR with DWI was available in 20 cases (10 boys), age range: 5 months-18 years (median 10.7 years). Tumors included: rhabdomyosarcoma (6), Ewing sarcoma (4), neurofibromatosis-1 (5), post-transplant lymphoproliferative disease (2), clear cell renal carcinoma (1), squamous cell carcinoma (1), and neuroblastoma (1). A total of 44 lesions were identified in session 1; 23 designated malignant and 21 benign. DWI did not detect any additional distant foci of disease in session 2. No change in final diagnosis or local extent of disease was noted in 19/20 cases. In one patient with Ewing sarcoma the local extent of osseous tumor was greater with DWI than with PET.

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: Purpose: 18 F-FDG Positron Emission Tomography (PET-CT) has been shown to be superior to other imaging modalities in assessment of soft tissue involvement with Langerhans Cell Histiocytosis (LCH) and is now commonly included in the staging workup of LCH. Our purpose is to analyze the utility of skeletal survey in addition to PET-CT scan for detecting bone lesions in children with LCH and to evaluate if skeletal survey can be eliminated from the staging workup of LCH in order to reduce radiation exposure.

This is a retrospective study of patients with biopsy-confirmed LCH diagnosed and treated at a tertiary care children's hospital between 2013 and 2018. The medical records, skeletal survey and PET-CT imaging records for staging and follow-up studies were reviewed. The distribution and number of LCH lesions on both skeletal survey and PET-CT were documented from our institution's PACS. Any reports with insufficient information or ambiguity were re-evaluated by a Pediatric Radiologist. Demographic information was recorded and the concordance between lesions seen on the skeletal survey and/or PET-CT scan was calculated. Results: Data from 10 children (60% female) with a median age of 4.6 years (range from 1 year 7 months to 15 years, 1 month) are reported. A total of 53 lesions in 40 different bones were evaluated. Forty-nine per cent of all lesions were identified on both PET-CT and skeletal survey. PET-CT identified an additional 34% of lesions not seen on skeletal survey, while skeletal survey identified 17% of lesions not seen on PET-CT. There were 9 calvarial lesions, all identified on skeletal survey while only 4/9 were identified on PET-CT scan. The median interval between the skeletal survey and PET-CT scan was 2.5 days (1-15 days). Conclusions: Skeletal survey continues to be important in the evaluation of LCH patients, especially when assessing lesions involving the calvarium. This may, in part, be due to the thickness of the PET-CT slices reducing lesion conspicuity and the marked avidity of the brain parenchyma on PET-CT obscuring lesions within the calvarium.

Poster #: SCI-064 mIBG optimized SPECT/CT improves interpretation and Curie score assignment Melissa C. Kong, James E. Potts, Helen R. Nadel, MD FRCPC, hnadel@stanford.edu; BC Children's Hospital, Vancouver, British Columbia, Canada

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: In the past, neuroblastoma patients frequently had mIBG and diagnostic CT scans performed at separate times. This practice sometimes caused issues in correlating findings from the two imaging modalities. A retrospective review of our entire experience aimed to confirm the added value of optimized co-registered contrast-enhanced diagnostic CT to I-123 mIBG SPECT/CT protocol in children with neuroblastoma. An additional objective was to identify if SPECT/CT improved Curie score assignment vs planar imaging.

We previously reported on a limited recent experience of optimized mIBG SPECT/CT. We have now completed a retrospective review of 384 SPECT/CT scans of pediatric neuroblastoma patients (aged 0 to 18 years) performed from April 2007 to June 2018. Added value of the co-registered CT was categorized as providing: increased sensitivity, improved localization, improved tumour delineation, improved differentiation between malignant and physiological or benign findings, detection of additional incidental findings, and/or no additional value when fused with the I-123 mIBG scintigraphy. Chi-square tests were performed to assess differences between diagnosis vs follow-up scans with added value from the co-registered CT. Curie score evaluation was performed using planar and SPECT/CT imaging. Results: Of all assessed scans, 70% of co-registered contrastenhanced diagnostic CT studies performed as part of the SPECT/CT scan provided added value (87% of staging scans at diagnosis (n=47), 67% in follow-up (n=337)). Overall, use of co-registered CT enhanced diagnostic sensitivity, tumour localization and delineation, and differentiation of malignant versus benign findings in many cases, and also identified additional incidental findings. Curie score assessment was improved in areas of equivocal lesions involving mainly the torso and particularly on follow-up examinations. Conclusions: Optimized mIBG SPECT/CT can reduce the need for additional imaging studies by improving accuracy of disease characterization and Curie score.

The value of physical and psychological distraction methods in reducing pain in paediatric nuclear medicine procedures. Purpose or Case Report: In paediatric nuclear medicine the majority of the scans require intravenous (IV) access to deliver the radiotracers. Children and parents often cite procedural pain as the most distressing part of their child's hospitalization. In our department, various pain management strategies including physical and psychological distraction methods and pharmacological intervention have been implemented for the reduction of procedural pain. The purpose of this study was to evaluate and compare different pain reduction strategies used in the paediatric Nuclear Medicine department.

The chart of 155 children (85 female) were reviewed retrospectively. Patients were categorized into 4 groups of 1. Maxiline (topical liposomal lidocaine) (n=17), 2. Pain Ease (vapocoolant) (n=71), 3. oral sucrose (n=48), and 4. no pharmacological intervention (n=19). Physical and psychological distraction were used in all patients. Therefore, group 4 only received physical and psychological strategies. Physical methods included supportive positioning, deep breathing, temperature considerations, massage Pressure or vibration and neonatal development strategies (e.g. nonnutritive sucking, facilitated tucking, swaddling, rocking). Psychological strategies included education, distraction with movies, books or storytelling, relaxation techniques. The pain perceived by the children after the IV access were compared in these 4 groups. Two types of pain assessment were used in this study: 1-Self reporting pain scale, and behavioural observational pain rating scale. Pain was reported on a scale 1 to 10. The average pain was also compared between patients who had one or two, and those who had more than two attempts for IV access. Results: The average pain score was relatively low in all 4 groups (Maxiline=2.8, pain ease=2.1, sucrose=2.7, and no pharmacology=3.4). There was no statistically significant difference between the 4 groups. In particular, although the average pain was slightly more in patients who didn't receive pharmacological intervention, it was not statistically significant. The average pain was 2.2 with one or two attempts, and 4.8 with of consolidation, nodule, and mass (solid versus cyst). Cyst was further evaluated for the presence of fluid, air, and internal membrane. The accuracy of MRI and MDCT for detecting pulmonary hydatid disease was obtained and compared between them. Interobserver agreement was measured with the kappa coefficient.

Results: Final diagnosis of pulmonary hydatid cyst was established in 25 children. Post-surgical histopathological confirmation was available in 12/25 patients (48%), while a positive hydatid serology was confirmed in 8/25 patients (32%). Remaining five patients (20%) were diagnosed to have pulmonary hydatid cyst based on the epidemiological setting and typical radiological findings. The accuracy of MRI and MDCT for detecting pulmonary hydatid cyst was 92.86%. There was no difference between MRI and MDCT for accurately detecting pulmonary hydatid cyst (p < 0.001).

Internal membranes were detected in 11/28 patients (39.28%) with MRI and 3/28 patients (10.71%) with MDCT. Almost perfect interobserver agreement was present between two independent reviewers (k = 1). Conclusions: Lung MRI without contrast is comparable to enhanced MDCT for accurately detecting lung cyst in pediatric patients with pulmonary hydatid disease. However, MRI provides an added diagnostic value by demonstrating internal membranes which is specific to pulmonary hydatid disease.

Combining chest radiographic findings and genomic scores to improve assessment of disease severity in pediatric community acquired pneumonia (CAP). Purpose or Case Report: CAP is a worldwide contributor to morbidity and mortality in children. Radiogenomics is an emerging specialty which correlates imaging features to gene expression to predict disease severity, therapeutic response and clinical outcomes. A genomic score termed Molecular Distance to Health (MDTH) is a biomarker that measures the global transcriptional perturbation in blood and has been shown to correlate with disease severity in pediatric CAP (Wallihan R, Front. Cell. Infect. Microbiol. 2018) . This study was undertaken to determine whether abnormalities on chest radiographs correlate with genomic and clinical markers of disease severity.

Initial chest radiographs of children (age 2 months -18 years) admitted to a single institution between February 1, 2011, and May 10, 2012 for CAP were reviewed retrospectively. An experienced pediatric radiologist blinded to the clinical data recorded abnormalities including peribronchial thickening, adenopathy, pleural effusion, and pulmonary opacities (characterized as band-like, streaky, consolidative, round or other). Clinical and laboratory data were collected including length of hospitalization (LOS), days of respiratory support (DOS), days of fever (DOF), blood culture, CBC, procalcitonin, C-reactive protein (CRP), and nasopharyneal/oropharyngeal swabs for viral and bacterial pathogen detection, and whole blood for transcriptional analysis and MDTH calculation. We used chest radiographic findings to classify patients according to clinical variables and the genomic MDTH score. We performed two-tailed unpaired t-test to compare groups stratified by radiological findings. Results: 144 chest radiographs were reviewed. Patients with pleural effusion (right or left) (n=43, 30%) showed significantly longer LOS, DOS, DOF, CRP and higher MDTH scores (all p <0.05) compared with those patients without pleural effusion (n=101). Likewise, patients with consolidative opacity (air space opacity) (n=82, 57%) showed significantly longer LOS, DOF, and higher MDTH scores (all p <0.05) compared with those patients without a consolidative opacity (n=62). Conclusions: In a cohort of children hospitalized with CAP, we identified chest radiographic findings that classified patients according to: 1) clinical markers of disease severity and 2) the MDTH genomic score. Combining radiographic and genomic markers should contribute to a more precise clinical disease severity classification in pediatric CAP. Purpose or Case Report: Superior mesenteric artery (SMA) syndrome, or Wilkie's syndrome, is an obscure condition in which an acute angulation between the aorta and SMA leads to compression of the third part of the duodenum. Subsequently, patients typically present with a constellation of gastrointestinal findings that closely resemble small bowel obstruction, early satiety, and anorexia. SMA syndrome is most often observed in the setting of rapid weight loss, wasting conditions, and corrective spinal surgeries, where a decrease in retroperitoneal fat diminishes the cushion between the aorta and SMA, causing vascular compression of the duodenum. Diagnosis of SMA syndrome is one of exclusion and is based on the combination of highly suspicious clinical findings and confirmatory diagnostic imaging evidence of obstruction. Upper GI Fluoroscopy can demonstrate a dilated first and second portion of the duodenum, with compression of the third portion, and delayed passage of contrast past midline. Ultrasound and computed tomography (CT)/magnetic resonance imaging (MRI) enable measurement of the aortomesenteric angle and distance, with normal values ranging from 25-60° and 10-28mm, respectively, and values indicative of SMA syndrome ranging from 6-15° and 2-8mm, respectively.We present a classic case of SMA syndrome in a 15-year-old patient. The patient presented with clinical symptoms consisting of scoliosis, chronic nausea and vomiting, weight loss, and anorexia. The patient then underwent Upper GI Fluoroscopy which demonstrated duodenal dilatation and delayed contrast transit past midline, achieved only after prolonged decubitus and prone positioning. Subsequently, abdominal ultrasound demonstrated an aortomesenteric angle of 10-18° and an aortomesenteric distance of 5 mm. Lastly, abdominal MRI demonstrated an aortomesenteric angle of 15° and an aortomesenteric distance of

A series of ten standardized PDSA testing sessions were completed in the clinical setting and included the use of a pre-screened ferrous-free person who transported a "control" projectile through the FMDS at separate intervals. The control consisted of either a 4.5" straight scissors or a 5" curved forceps divided into two groups; exposed and non-exposed. All controls were deemed a "projectile hazard", according to the ASTM deflection test (Ref) . Results: Significant gaps in effectiveness and programmatic variables were identified within the expected performance of the FMDS installed. With the current settings and modifications made, we discovered a gap that exists at the center region of each door passageway where detection was minimal.Through our project, 42 new practice changes were implemented and 68 existing process improvements were achieved. A 78% reduction in alarm rates was achieved and a 100% reduction of incidents where hazardous projectiles entered zone IV was realized. These reductions were achieved via optimization and customization of the latest FMDS technology and various process changes and improvements.

Institutions where these devices have been installed may not be able to reliably detect metallic objects classified as projectile hazards. Validation of installed systems can and should be accomplished in order to optimize the level of sensitivity and effectiveness of each FMDS installed in situ. Purpose or Case Report: To describe and discuss some of the challenges of performing DEXA scans on pediatric patients including positioning, technical obstacles, reference data and post-processing and describing how to overcome some of these challenges.1. Positioning: Many patients with syndromes, have scoliosis, para or quadriplegia, or contractures that make placing the patient on the table difficult. In these patients, imaging has to be adapted to the patients' abilities and some components may need to be eliminated. With cerebral palsy and muscular dystrophy for example, the whole body and AP spine might have to be deferred and only a hip and forearm obtained.2. Technical obstacles: There are many technical obstacles that can present challenges when performing DEXA scans on pediatric patients. Patients who have prostheses or metal rods present technical obstacles. There is no way to remove the metal artifacts in a whole body scan on these patients, so a hip or forearm might be the most accurate way to obtain the patient's bone density due to this technical factor.3. Reference Data: The reference data for pediatrics is limited in national data bases. For example, a total Z-score will not be factored for children under the age of 5 because there is not enough information in the national database for comparison. There is also not enough information on children of certain ethnicities. This presents a problem when diagnosing and treating children with abnormal bone density.4. Post-processing could be different for each user. It can also be different for machines manufactured by different companies. For example, a machine made by Hologic might produce different numbers than a machine made by General Electric. Also, if the user does not place the post-processing tools the exact same way the prior user did, the results can vary.Overcoming some of these challenges has presented opportunities to grow. Sedation is used for children who may not be able to be perfectly still for a Dexa and positioning tools help on some of the more challenging patients. Learning how to work around prostheses and working with the ordering physicians and radiologists with the limited amount of reference data are a few ways of growing from these challenges. Placement of post-processing tools are the key to providing key information in comparing Dexa scans for patients who have them regularly for evaluations. Staff Education is the largest challenge to overcome. Purpose or Case Report: PET/MRI is a relatively new imaging modality whose efficacy is still being determined. One of the major draws to PET/MR over PET/CT is the reduction in radiation exposure to patients. This is especially desirable in the pediatric population due to the likelihood of multiple exposures during their lifetime, and the increased sensitivity they have to radiation. A typical whole body PET/CT exam can take about 30 minutes, whereas a typical whole body PET/MR exam takes about 90 minutes at our institution. The introduction of a new 3D T2 Dixon technique sequence for PET/MR has the potential of decreasing total scan time significantly, however maintaining current image quality and diagnostic value is critical. Our objective is to test out this new sequence to see whether scan times are reduced and if it is a viable diagnostic replacement for our current T2 sequence.

Discussion over speed stitching in the pediatric world Amal Baida, amal.baida@choa.org; Stephen Simoneaux; Children's Healthcare of Atlanta, Duluth, GA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: The purpose of this study is to outline the benefit of speed stitching and recognize it as one of the best stitching tools available for the pediatric population. Looking at the history of stitching, we started by using the CR for scoliosis and long bones stitching. The time of exposure, the radiation dose, and image quality were not optimal. Moving to automatic DR stitching was a good step in the right direction. However, there was still a frustration from having to repeat so many exams because of the high possibility of motion captured between exporsures. That's when speed stich came to play with fast acquisation and less operator interference while doing the exam.

Recognizing the challenges of doing scoliosis or long film stitching in the pediatric world led us to start this conversation. Motion and exposure factors were big topics.CR's low bone resolution has been posing a challenge with the risk of overexposure; DR's auto stitching brought the fast image acquisition with better quality and high volume capacity, still posing some issues with manual stitching and the overlap of vertebras. Discussions were conducted on several scoliosis and example cases were identified. Most importantly, the accuracy of the images for optimum diagnosis was our priority at CHOA. Speed stitching was identified as a solution for many of the inefficiencies in other technologies. With Speed stitching, we were able to reduce a scoliosis exam from 15 minutes to about 4 minutes with lower radiation and better quality.

Results: A case-based education for speed stitch vs other ways of stitching was developed and comparison cases with other methods of stitching were demonstrating for comparison and contrast. Reviews by the radiologists determine the advantages of speed stich and full transition to that technology is in process for the whole system at CHOA. Conclusions: Beside the higher initial cost, using speed stitch was faster. It demonstrated better image quality, less dose and better accuracy.Technologists moving from CR/DR to speed stitch had no challenges or issues learning the new technology. Mostly everyone thought it was more efficient and easier to use. Purpose or Case Report: In 2017, roughly 2 trillion (2,000,000,000,000) medical images were produced, reviewed, reported, archived, and used in the detection and management of disease. Based on historical trends, this number has doubled every 5 years and is accelerating. This explosive growth in imaging data has created major opportunities for the use of Artificial Intelligence (AI). The question is less whether radiologists, and technologists, will be replaced by AI (they will not) and more about whether we could survive without AI. Although intelligent algorithms have been used for some time in segments of the imaging field, new methods of machine learning, based particularly on "deep learning", are much more powerful. Many of the deep learning publications today point to the promise of significant advances in efficiency, precision, reproducibility, and prognostic abilities.If AI will not replace radiologists/technologists but rather augment them with tools to meet the rising demands for diagnostic imaging, then it is imperative that we have a basic understanding of the concepts and language that defines this area of knowledge. In the not so distant past the average technologist understood the basics of film processing but wouldn't even recognize the words DICOM or EMR; we are now at that point of change with AI. Deep learning, machine learning, neural networks, ground truth, the list goes on. The goal of this presentation is to provide a basic framework of the concepts, terminology, and references to how AI has, and likely, will be employed in medical imaging, thus making us better practitioners and partners with this technology. Purpose or Case Report: Langerhans Cell Histiocytosis (LCH), is a disorder that primarily affects children, but can affect individuals of all ages. Langerhans cells are cells that are responsible for regulating immune system in our bodies. They are mostly found in the skin, spleen, lymph nodes, liver and bone marrow. In patients who have LCH, these cells grow and multiply excessively. The abnormal growth of the Langerhans cells causes a formation of tumors called granulomas. LCH can affect different areas of the body: skin, nails, lymph nodes, gastrointestinal tract, central nervous system, pituitary and thyroid gland, liver, lungs and bones. The severity and symptoms of the disease vary in individual patients and are dependent on the organs and systems affected. Oftentimes, LCH can be found in multiple areas of the body and when that happens, the disease becomes a multisystem disease. The most common system affected by LCH, seen in about 80 % of individuals affected, is the skeletal system. Granulomas, which develop most commonly in the flat bones such as skull and long bones of arms and legs, cause sclerotic and lytic lesions that can in turn become the cause of pathologic fractures. Therefore it is crucial, to recognize the radiographic signs of skeletal manifestations of LCH. Radiography is the preliminary imaging of choice and skeletal surveys are oftentimes the best assessment of the status of LCH prior and post treatment. The purpose of this abstract is to describe radiographic appearances associated with Langerhans Cell

Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.

Purpose or Case Report: To compare the radiation doses that result from different digital radiography imaging methods on neonatal patients in Giraffe Beds using phantom imaging and radiation measurement devices. The lowest dose options will be considered to design an imaging process.

Using technical factors of 0.63 mAs and 60 kVp as a constant, test images were taken on a phantom in a Giraffe Bed using the following methods:1) Cassette placed in the bed tray and image taken with the hood down at 40" SID (Source-to-Image Distance)2) Cassette placed directly under phantom on bed mattress (not in tray) and image taken with the hood down at 40" SID3) Cassette placed in the bed tray and image taken with the hood up at 36.5" SID4) Cassette placed directly under phantom on bed mattress (not in tray) and image taken with the hood up at 32" SID5) Cassette placed directly under phantom on bed mattress (not in tray) and image taken with the hood down at 32" SIDFor each imaging method, the radiation dose and EI (Exposure Index) number were recorded. EIThe radiolucent bed hood caused a decrease of approximately 13% of the radiation reaching the image detector. Utilizing the imaging tray in the Giraffe Bed caused a decrease of approximately 5% of the radiation reaching the image detector. Images taken at the shortest SID resulted in the highest doses and least optimal EI numbers as anticipated. Please note Image 1 had an inconsistant EI measurement. Conclusions: During the study, varying the SID had the most significant impact on radiation doses and EI numbers compared to the other factors caused by the Giraffe Bed. The maximum SID allowed when the bed hood is raised is only 32" which produces higher radiation doses and EI numbers compared to imaging through the bed hood in the down position at a greater SID. When designing an imaging process, the primary consideration should compensate for a greater SID to reduce the radiation dose and EI number. Imaging at a greater SID also allows the technologist to select a wider range of technical factors and avoid over exposure of the smallest neonatal patients. Purpose or Case Report: Anesthesiologists face challenges in practicing in locations outside of the operating room. In particular when working in the MRI environment there may be unfamiliarity of the physiologic monitors and other equipment, MRI safety policies and procedures, limited access to expected equipment or additional anesthesia providers, unsure imaging protocols and positions as well as inability to rescue patients in the MRI scanner room.

Radiology must also understand how lengthy or unpredictable imaging times, positioning, and monitoring and equipment issues challenges anesthesia providers as they care for patients in the MR environment.Therefore anesthesiology and radiology must recognize possible risk factors while caring for patients in MRI and they must familiarize themselves with policies and procedures of both departments to ensure safe practice for the anesthetized children.

To understand the risk of both anesthetizing patients and the risk MR safety, radiology and anesthesiology must partner to form a structure where anesthesiology, radiology and nursing collaborate to ensure that the needs for patient safety and the needs of MRI are met but ensuing that the priorities are first to the patient, next to the enterprise and third to the individual departments and division in order to provide the safest environment for our patients.The structure has two committees:Radiology, Anesthesia and Sedation (RAS) Committee and is an active working group to improve patient care and flow with ongoing initiatives that utilize appropriate metrics to measure improvement. Issues that cannot be solved in committee are escalated to the Operational Governance Oversight Committee.Operational Governance Oversight Committee (meets quarterly): Provides oversight to ensure that the needs of all constituents are considered and met to the best degree possible. Overriding priorities are first to the patient, next to the enterprise, and third to the individual departments and divisionsChairs from Anesthesiology, Pediatrics, Radiology, Chief Operating Officer, Chief Nursing Officer Conclusions: The MRI environment can be challenging to anesthesiologists as they care for patients outside of their usual environment. Collaboration between both radiology and anesthesiology can improve the experience and safety of anesthetized patient during MRI exams. 

MD 2010 Janet L. Strife, MD 2011 Carol M. Rumack

FACR 2019 Marta Hernanz-Schulman

Angkor Hospital for Children

MBchB, Kenyatta National Hospital

Legg Perthes Disease: The Scientific Basis for the Various Methods of Treatment and Their Indications

Wrap MRU

Cramer BC, Walsh EA. 2001 -Aneurysmal Bone Cysts In Children: Percutaneous Sclerosing Therapy, An Alternative To Surgery

-MR Imaging of Kidneys: Functional Evaluation Using F-15 Perfusion Imaging

Differential Regurgitation in Branch Pulmonary Arteries after TOF Repair

Diagnostic Imaging, Texas Children's Hospital, Houston, TX 2005 -Evaluation of High Resolution Cervical Spine CT In 529 Cases of Pediatric Trauma: Value Versus Radiation Exposure

Damien Grattan-Smith, MBBS, Department of Radiology, Children's Healthcare of Atlanta

Cincinnati Children's Hospital Medical Center 2009 -Improving Patient Safety: Effects of a Safety Program on Performance and Culture in a Department of Radiology at a Children's Hospital -Lane Donnelly, Cincinnati Children's Hospital Medical Center

Jonathan Rubin 2015 -Contrast Enhanced Ultrasound in the Assessment of Pediatric Solid Tumor Response to Anti-Angiogenic Therapy; Beth McCarville, MD, Department of Radiological Sciences, Division of Diagnostic Imaging, St. Jude Children's Research Hospital

Association of T2 Lymphatic Imaging in Single Ventricle Patients After Superior Cavopulmonary Connection with Acute Post-Fontan Outcomes; David Biko, MD, Department of Radiology, The Children's Hospital of Philadelphia

Imaging in the Diagnosis of Experimental Pyelonephritis in Piglets

BG 2001-Mitochondrial Disorders Of Oxidative Phosphorylation In Children: Patterns Of Disease Palasis S

Volumetric Localization of Somatosensory Cortex in Children Using Synthetic Aperture Magnetrometry. Xiang J, MD, PhD, The Hospital for Sick Children

Outstanding Basic Science Research Poster -Imaging of the Diaphragm in Neonates and Young Infants, with Special Emphasis on Diaphragmatic Motion. Epelman M, Navarro O, Miller S Department of Diagnostic Imaging, Hospital for Sick Children

MD 2015-Scientific Poster: Infant Bone Age Estimation Based on Fibular Shaft Length: A Validation Study, Andy Tsai, MD, PhD, Boston Children's Hospital

Children's Hospital, Cincinnati, OH 2016-Educational Poster: Pediatric Thyroid Cancer: Common Sonographic Appearances and Pitfalls; Claudia Martinez-Rios, MD, Diagnostic Imaging, The Hospital for Sick Children

MD 2017-Case Report: Cutaneous Metastases of Infantile Choriocarcinoma can Mimic Infantile Hemangioma both Clinically and Radiographically; Logan Dance, MD; Patricia Cornejo, MD; Mittun Patel, MD, Phoenix Children's Hospital, Phoenix, AZ 2017-Educational Poster: Sonographic Evaluation of Diaphragmatic Motion: A Practical Guide to Performance and Interpretation

2017-Educational Poster: Nuts and Bolts: A Radiologist's Guide to Orthopedic Hardware Utilized in the Lower Extremities of Children

MD 2017-Scientific Poster: Sinusoidal Obstruction Syndrome Causes Increased Liver Stiffness

PhD indication for posterior line placement, type of catheter, number of catheter days, rates and reasons for replacement, and infections. Mechanical complications were also assessed

Children's Hospital of Chicago

Dynamic Contrast-Enhanced MRI Using a 3D Radial Acquisition: Potential Applications in Musculoskeletal and Bone Marrow Assessments Mitchell Rees

Houchun Hu, PhD; Radiology, Nationwide Children's Hospital

Poster EDU-007 (T)

Poster EDU-002 (T)

Fiama Paper

Poster SCI-001 (T)

Poster EDU-002 (T)

Marjolaine Paper

Poster EDU-013 (T)

Amer Poster SCI-065

Bhupender Paper

A Abdomen Poster

Abusive Head Trauma Paper 125

Acetabular Fractures Poster EDU-055

Acquired lesions Poster EDU-064

Acute scrotal pain Poster CR-003 (T) Adnexal torsion Paper 022

Ambiguous genitalia Poster EDU-011

Appropriateness Poster SCI-006

Appropriateness Criteria Poster EDU-001 (T)

Brachial plexus birth injury (BPBI) Poster EDU-015 (T)

Cardiac Iron Paper 123

Melanocytosis Poster

Mullerian Duct Anomalies Poster EDU-034

Multimodality guidance Poster SCI-039

Neutropenic Fever Poster SCI-056

Newborn Poster EDU-004 (T)

Ossified soft tissue Poster EDU-052

Osteoid Osteoma Paper 010

Pediatric Hip Poster

Pediatric Imaging Poster

Pediatric oncologic emergency Poster EDU-084

Pediatric scrotal Ultrasound Poster CR-003 (T)

Radiogenomics Poster SCI-069

Radiographic findings Poster EDU-012 (T) Radiography Poster

Seizure Poster

Septic Emboli Poster

Sinus Mucosal Thickening Poster SCI-056

Prof. Malik's research group has worked on many different topics in computer vision, computational modeling of human vision, computer graphics and the analysis of biological images. Several well-known concepts and algorithms arose in this research, such as anisotropic diffusion, normalized cuts, high dynamic range imaging, shape contexts and R-CNN. He has mentored more than 60 PhD students and postdoctoral fellows. His publications have received numerous best paper awards, including five test of time awards -the Longuet-Higgins Prize for papers published at CVPR (twice) and the Helmholtz Prize for papers published at ICCV (three times). He received the 2013 IEEE PAMI-TC Distinguished Researcher in Computer Vision Award, the 2014 K.S. Fu Prize from the International Association of Pattern Recognition, the 2016 ACM-AAAI Allen Newell Award, and the 2018 IJCAI Award for Research Excellence in AI. He is a fellow of the IEEE and the ACM. He is a member of the National Academy of Engineering and the National Academy of Sciences, and a fellow of the American Academy of Arts and Sciences.

Results: After reviewing the exhibit, the reader will be aware of the spectrum of chILD in infants, and be able to use the discussed imaging based algorithm to assist in efficient and accurate diagnosis of various chILD entities in the infant population. Conclusions: chILD in infants constitutes a diverse group of lung abnormalities which can be complex and challenging to diagnose. Aside from the infants with diffuse development disorders, whom typically are only imaged with chest radiographs, the remainder of the diseases in the chILD spectrum presenting in infants can be approached with a CT algorithm utilizing the stepwise assessment of lung volumes, ground glass, and cysts. In conjunction with the patient's demographics and clinical presentation, this algorithm can aid the radiologist in making an accurate and timely diagnosis. Purpose or Case Report: BPM's are often identified prenatally; while some have more detailed imaging and description, many are loosely called congenital pulmonary airway malformations (CPAM). A chest radiograph is usually obtained at birth, but CT imaging is often deferred until 3-6months of age in asymptomatic babies, when surgical removal is being considered. Participation in a presurgical conference has indicated that there is poor standardization of both performance and interpretation of CT for BPM's. High quality studies are most often hampered by poor timing of imaging, poor vascular opacification and obscuration of pathology due to atelectasis. There are four key internal feature of BPM's that help with lesion characterization, differential diagnosis and management decisions. These include: systemic arterial supply; bronchial mucoid impaction; overinflated lung and macroscopic cysts. Reliable recognition and description of these features in all cases is essential for guiding surgical decisions since some lesions can be treated conservatively, especially those with just hyperinflation and mucoid impaction. A feature that tends to be overlooked is mucoid impaction, indicative of bronchial atresia. Cystic changes and overinflated lung may be mischaracterized. Small systemic arteries can be missed. Multiplanar reconstructions and interaction with maximum intensity projections and a 3D dataset are very helpful. This poster aims to educate by providing multiple illustrative imaging examples and a standardized report template useful for radiologists, clinicians, research and registries.

There is MAGEC going on in Ultrasound Amy Winer,amy.weiner@cchmc.org; Ultrasound, Cincinnati Children's Hospital, Cincinnati, OH Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: 1.To identify what are MAGEC rods and how ultrasound is utilized in the lengthening procedure to reduce the number of x-rays that they are exposed to.2. To describe the use of Ultrasound to aid with lengthening the MAGEC rods in Scoliosis patients.3. Decscribe the pros and cons of Ultrasound use.

Scoliosis is a sideways bending or curvature of the spine. At times, surgical intervention is required to help correct or stop the curvature from getting worse. By using the MAGEC rods, this can reduce the number of surgeries and radiation from xrays that a patient is exposed to.Ultrasound is used to help measure the growth of the growing rods at each lengthening and reduce the number of xrays that the patient is getting. In this poster, I will discuss the procedure of performing an lengthening of the MAGEC rods and the benefits and downfalls of ultrasound. Results: What is scoliosisTreatment of scoliosisMAGEC rodswhat is it?How do they workAnatomy of the rodsUsing Ultrasound as a measuremen toolBenefit of Ultrasound usageDownfalls of Ultrasound usage Conclusions: Using Ultrasound to help with the lengthening MAGEC rods can save a patient from the exposure or xrays along with reducing the number of surgeries that a patient will need in their lifetime. Purpose or Case Report: With an incidence of 3.25%, breast masses in the pediatric population are a relatively rare phenomenon. Despite this, breast masses are a substantial source of anxiety and concern for parents and patients alike, largely due to the increased awareness of breast cancer in the adult population. Fortunately, the vast majority of masses are benign, and pediatric breast malignancies constitute less than 1% of all pediatric malignancies. Moreover, malignancy tends to be secondary to metastatic disease from lymphoma, leukemia, or rhabdomyosarcoma, as primary breast carcinoma is exceedingly rare. Although initial sonographic characteristics of breast masses may be nonspecific, recommendation for further evaluation with biopsy and/or excision of the mass is usually not recommended due to the rarity of malignancy, and avoidance of disrupting immature breast parenchymal tissue. In adults, the Breast Imaging-Reporting and Data System classification is quite accurate for dictating management recommendations. However, this system grossly overemphasizes the risk of malignancy in pediatric patients, as imaging findings are usually discordant with histology. Currently, there are no standardized guidelines for management recommendations of pediatric breast masses, and short-term follow-up ultrasound is usually recommended to evaluate for malignant potential. The purpose of this educational exhibit is to compare the sonographic abnormalities of breast pathologies arising from normal breast development, including but not limited to gynecomastia, mastitis, and abscesses, from those arising from neoplastic processes including fibroadenomas, hemangiomas, arterio-venous malformations, and phyllodes tumors. In addition, the current literature on management recommendations, including indications for MRI and biopsy/excision of breast masses, will be reviewed. Finally, this exhibit will discuss the important role pediatric radiologists play in understanding the epidemiology and natural history of breast pathologies, enabling accurate characterization of masses and appropriate treatment recommendations to further guide patient management. Purpose or Case Report: Pleuropulmonary blastoma (PPB) is a rare, intrathoracic, malignant tumor in the pediatric population with approximately 500 cases reported worldwide. Over 90% of these cases are in patients below the age of 6. The spectral morphology of PPB is used to classify the lesions along a continuum from the least malignant to the most malignant: type I (cystic) 14%, type II (solid and cystic) 48%, and type III (solid) 38%. Congenital lung cysts are not known to degenerate to become PPB, but the cystic type I PPB may progress to the more aggressive type II or type III PPB. In addition, PPB is associated with cystic nephroma in 30% of cases, and has been linked to the DICER1 mutation which puts these patients at risk for other tumors. For example, the genetic basis of the PPB familial syndrome (which is the heterozygous loss of function mutation of DICER1) includes PPB, cystic nephroma, ovarian Sertoli-Leydig cell tumors, ciliary body medulloepithelioma, nodular hyperplasia and differentiated carcinoma of the thyroid gland, pituitary blastoma, pineoblastoma, nasal chondromesenchymal hamartoma, and cervical embryonal rhabdomyosarcoma.The purpose of this educational report is to demonstrate various presentations and identify distinguishing features of each type of PPB as seen on initial radiographs with correlation on subsequent CT scans. Only cases where the PPB and type were confirmed by pathology are included. At least one case of the PPB familial syndrome will also be presented. Positive DICER1 mutations will be provided when available, as this information is increasingly used to aid in the treatment decisions. Early recognition of PPB with timely investigation for cystic nephroma and DICER1 mutations can lead to improved patient outcomes. Purpose or Case Report: In 2017, children under the age of 15 accounted for only 10% of the 10 million Mycobacterium tuberculosis (TB) infections estimated by the World Health Organization. Child carriers of TB pose health risks to their adult caregivers, and children can exhibit greater susceptibility to significant health risks from the infection. Nontuberculous mycobacterial (NTM) infection presents disparate health risks and can generate imaging specific findings. This presentation addresses radiologic manifestations of mycobacterial disease as a means to educate pediatric radiologists given the medical significance of mycobacterial infection.

This exhibit will discuss how children may manifest TB infections differently than adults and describe imaging findings of pulmonary and extra-thoracic TB infection in children. Contrasting findings in multi-system NTM infection will also be detailed. The presentation will provide radiologic manifestations of verified mycobacterial infection and emphasize unique characteristics of mycobacterial infection. Results: Pulmonary TB in children may present with diffuse pulmonary disease and/or pleural effusions that can appear more impressive than the clinical presentation seemingly suggests. Extra-pulmonary TB can cause brain abscesses, osteomyelitis, and intra-abdominal infection which have distinct appearances compared to typical bacterial infections. Pathogenic NTM organisms include Mycobacterium avium-intracellulare, Mycobacterium kansasii, Mycobacterium xenopi, Mycobacterium fortuitum, and Mycobacterium chelonae. MTB commonly infect the soft tissues and cervical lymph nodes as well as the lungs. The radiologic and clinical presentation of mycobacterial infection often depends on the immune status of the child. Mycobacterial recovery is often complicated by inability to obtain a sufficient biologic specimen and lack of growth on routine contrast media. For this reason, pediatric radiologists play a key role in raising appropriate managementguiding concerns for disease. Conclusions: Mycobacterial infection can present differently in children. Radiographic findings suggestive of pediatric TB should prompt testing in relevant contacts.

Big bubbles, little bubbles, bubbles everywhere: A review of macro and microcystic lymphatic malformations in less common anatomical locations.Marian Gaballah, D.O., mgaballah@northwell.edu; Rachelle Goldfisher, MD; Zucker School of Medicine at Hofstra/Northwell, New Hyde Park, NY Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Lymphatic malformations (LMs) are low-flow vascular malformations which are composed of dilated lymphatic channels, forming septated cyst-like structures (2). LMs are the second most common type of vascular malformation, second to venous malformations (1). The most common locations are in the neck, followed by the axillary region. On MRI, LMs are multiloculated, T2 hyperintense lesions, which may have fluid-fluid levels, and are without flow voids (1, 2). They can involve multiple tissue planes and do not regard anatomical and fascial boundaries (2). Cystic lymphatic malformations are further divided into microcystic, macrocystic, or mixed, based on the size of their cystic components. Macrocystic LMs are composed of larger cysts, while microcystic LMs are composed of smaller cysts and may appear solid on imaging.We present ten cases of microcystic, macrocystic, and mixed lymphatic malformations in a variety of anatomical locations. In addition to demonstrating the imaging findings, we present a review of the literature in regards to each anatomical region. Lymphatic malformations in this presentation include right orbit (n=1), mediastinum (n=2), pulmonary bronchovascular bundles/pleural space (n=1), retroperitoneum (n=1), mesentery (n=2), perirectal and scrotal (n=1), lower extremity (n=2). 6 of these children also had additional sites of T2 hyperintense disease involving the bones or spleen, suggestive of additional lymphangiomas.This presentation summarizes ten cases of lymphatic malformations in a variety of less common anatomical locations and a review of the pertinent literature Purpose or Case Report: Balancing PEEP settings in mechanically ventilated premature infants to maximize air exchange while minimizing barotrauma is typically evaluated using bronchography. A 4D dynamic airway computerized tomography protocol (4D CT) was developed as a less invasive method to evaluate large and small airways collapse at variable PEEP settings. Our purpose was to evaluate the radiation dose from 4D CT and to demonstrate the number of cycles impact on the radiation dose. Methods & Materials: 39 pediatric patients <5 years who underwent imaging as part of routine clinical care between 1/1/16-10/1/18. Dynamic sequential CT scan was performed on Siemens Flash, fixed 70 kVp, fixed 10 mAs. One 4D CT imaging cycle was defined as the time required for one full inspiratory-expiratory cycle at a given PEEP, which ranged between 2-8 seconds. Each cycle contained 7 rotations. The body and skin effective radiation doses were estimated using two methods: (1) the body effective dose is calculated using Monte Carlo simulations of a library of male and female anthropomorphic size and age-specific phantoms. (2) NanoDots optically stimulated luminescent dosimeters were used to measure the peak skin dose (PSD) by placing them on the chest of a newborn phantom for a different number of cycles. Results: Increasing the number of cycle/rotations increased the PSD, dose length product (DLP), effective dose, and the lung organ dose. The estimated effective dose varied depending on the patient's DLP value, patient's gender, and age weighting factors. The average effective dose was higher for female patients. From the simulation's results, the highest organ dose was received by the lungs. The average value for lung dose was 0.9 mSv. Conclusions: Irradiation over the same anatomic region by 4D CT results in accumulation of radiation dose and raising the concern for the potential deterministic effect of skin injury. The maximum measured entrance skin exposure by the 4D CT was orders of magnitude lower than the threshold dose for early transient erythema (2000 mGy). As the PSD increased by increasing the number of rotations, reducing the number of imaging cycles may reduce the overall patient radiation exposure. A differential effective dose in female patients is due to higher end-organ risk from radiation scatter. Extra caution to limit cycles should be exercised when using 4D CT technique in children under 1 year. In addition, DLP would be a better dose metric than CTDI because of DLP metric account for the total irradiated area. Purpose or Case Report: Delayed phase imaging increases the sensitivity of detection of injuries to the urinary tract and also assists in characterizing solid visceral organ injuries at the expense of doubling the radiation dose to the patient. If institutions can lower the rate of these examinations, the cumulative radiation exposure reduction would be substantial. Here we evaluate the rate of delayed phase imaging in the pediatric trauma population at our Level I trauma center as well as the frequency with which these patients demonstrate an indication on portal venous imaging for delayed phase acquisition. Finally, there are minimal guidelines dictating the appropriateness of delayed imaging in pediatric trauma. We include a review of the literature in order to elucidate appropriate indications and help guide clinicians to make evidenced based decisions.

A retrospective chart review was performed analyzing data of pediatric (0-18 years) trauma activations at our institution with a CT chest/abdomen/pelvis, between January 1 2016-January 1 2018. The primary variable analyzed was acqusition of delayed imaging. Indications for delayed imaging on portal venous phase imaging, including solid organ injury, pelvic fracture, and free fluid, were also reviewed. Given the potential for physiologic free fluid in postpubescent girls, data was also analyzed excluding free fluid as an indication in girls older than 12 years old. Results: 134 patients met the inclusion criteria. Delayed imaging was acquired in the majority of patients (91%, n=122). There was a near even split for presence/abscence of an indication for delayed imaging on the portal venous phase imaging. There was no statistical correlation with an indication for delayed imaging and the acquisition of delayed imaging (p=0.475). When accounting for each indication independently, there was no statistical correlation between a specific indication and acquisition of delayed imaging. Conclusions: We found no correlation between indication for delayed-phase imaging and its acquisition in pediatric trauma patients at our Level I trauma center. Delayed phase imaging was obtained in the vast majority of these patients, irrespective of indication. Improving referral education and adoption of a protocol in which delayed imaging is only obtained in pediatric patients with appropriate indications could reduce the effective radiation dose to the pediatric trauma patients at our instition by 25%. Review of the literature corroborates these findings. Purpose or Case Report: Given the relative sensitivity of pediatric patients to radiation, the need to limit exposure to the region of interest is of the utmost importance in pediatric radiography. This study assesses the practice of acquiring pediatric chest radiographs at a community hospital in which the imaging contract was acquired by a radiology practice with subspecialty-certified pediatric radiologists. Pediatric radiologists in the new radiology practice saw a need for improved collimation of pediatric chest radiographs at the community hospital. Many radiographs exhibited poor collimation and included much of the abdomen. This study examines an initiative to measure and improve radiograph quality. Methods & Materials: All pediatric chest radiographs (n = 50, average age 5.6 years) obtained at the community hospital during 4 consecutive weeks were reviewed. The following Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Congenital high airway obstruction syndrome (CHAOS) is a rare life-threatening disease and prenatal diagnosis is essential. Some characteristic features seen on fetal MRI are well-known: primary lesion (upper airway obstruction) and secondary changes (dilated trachea, flattened/inverted diaphragm, enlarged and hyperintense lung). It is also reported that these secondary changes may be reduced if tracheoesophageal fistula (TEF) is present. The aim of this study is to evaluate the accuracy of fetal MRI in the prenatal diagnosis of CHAOS by comparing results with those of normal fetuses.

The MRI images from eight fetuses with CHAOS (29±6 weeks' gestation [mean±SD] ) and 37 fetuses with no thorax abnormalities (32±2 weeks) were assessed retrospectively. The fetuses with CHAOS were selected from among those who were diagnosed at our institution from 2006 to 2018 (four CHAOS fetuses with TEF were also included), and the normal fetuses were selected from consecutive fetal MRI performed in 2017 and 2018. Identification of the upper airway was evaluated in both groups (false negative rate [FNR] in the CHAOS group and false positive rate [FPR] in the control group). Measurement of tracheal diameter (TD), craniocaudal/antero-posterior ratio in the right diaphragm (CC/APR), cardiothoracic ratio (CTR), and lung-to-liver signal intensity ratio (LLSIR) were also carried out in both groups. For comparison between the CHAOS group and the control group, a t-test was used. Also, CHAOS fetuses with TEF were evaluated in the same way as described above. Results: Upper airway obstruction could be detected in all fetuses with CHAOS (FNR=0%), while the upper airway could not be clearly identified in five fetuses in the normal group (FPR=13.5%). There was no statistical difference in TD (3.6±1.0 mm, 3.2 mm±0.5 mm, p=0.13) and LLSIR (2.64±0.68, 2.54±0.66, p=0.35) between the CHAOS group and the control group. However, CC/APR (4.4±14.1%) and CTR (45.9±7.2%) in the CHAOS group were significantly lower than those of the control group ( Purpose or Case Report: Congenital pulmonary airway malformations (CPAM), bronchopulmonary sequestrations (BPS), and hybrid lesions are the most common congenital lung lesions. They are primarily diagnosed prenatally via ultrasound and further characterized by MRI. While most affected neonates are asymptomatic at birth, some may experience varying severities of respiratory distress requiring intervention. We seek to develop a prognostic model for prediction of post-natal outcomes in patients with congenital lung lesions using fetal MRI calculated observed to expected normal lung volume (O/E NLV).

We did an IRB approved, retrospective study of patients with congenital lung lesions who underwent fetal MRI and received pre-and post-natal care at our institution from 2006 to 2016. 163 cases were referred to our institution for prenatal diagnosis of CPAM, BPS, or hybrid lesion. 68 of these patients had prenatal MRI performed at our institution. 8 patients were excluded due to "disappearing" lesion at the time of MRI or non diagnostic study. Statistical analysis was performed using Chi-square and Student's t-test. MRI was reviewed by 2 pediatric radiologists to determine volume of Purpose or Case Report: To share these uncommon but important cases of Midgut Volvulus (MGV) without malrotation.

-15-day-old Male presented with bilious emesis. An Ultrasound was perfomed.-9-year-old Female with History of cerebral palsy, Gastro-jejunostomy tube fed. Presented with abdominal pain and hematemesis. A CT scan was obtained. Results: -The 15-day-old patient Ultrasound showed: SMA Cut-off as well as Superior mesenteric vein (SMV) Cut-off. In addition, congestion of the mesenteric venous vasculature and a subtle swirling of the bowel around the SMA. The diagnosis of Midgut Volvulus was made. No UGI examination was performed. At Surgery, MGV without malrotation were found.-The 9-year-old CT showed: SMA Cut-off as well as (SMV) Cut-off. In addition, a subtle swirling/kinking of the bowel around the SMA. The diagnosis of Midgut Volvulus was made. No UGI examination was performed. At Surgery, MGV without malrotation were found in addition to midgut infarction. Conclusions: These 2 cases illustrate the fact that a typical MGV can occur in the absence of malrotation. Although, no UGI was performed, we speculate that UGI examination to the ligament of Treitz may have been negative in these cases since there was no malrotation at surgery.SMA cut-off sign was important to make the diagnosis at Ultrasound and CT. Purpose or Case Report: To apply the Couinaud' system of segmentation and PRETEXT staging system of the liver to the tumor staging, combined with MSCT for evaluate the effectiveness in pre or post therapy of hepatoblatoma.

We retrospectively reviewed the outcomes of 67 cases that were diagnosed as hepatoblastoma in our hospital between July of 2012 and September of 2017. The patient ages ranged from 4 months and 5 years old (median age 3.2 years old). Couinaud' system of segmentation of the liver and PRETEXT staging system in MSCT were used to evaluate hepatoblastoma. According to the staging system results and imaging features, received the therapy relatively. We evaluate the patient's response to the therapy of these 67 cases. All cases were confirmed by pathology, diagnostic accuracy was calculated.Results: Among the 67 cases: (1) The density of tumors were heterogeneous and lower than hepatic tissues, 22/67 had calcifications.(2) MSCT PRETEXT staging results: Stage I 7 cases, Stage II 31 cases, Stage III 26 cases, Stage IV 3 cases; PRETEXT staging was used again according to the results of operation: Stage I 7 cases, Stage II 27 cases，Stage III 31 case, Stage IV 2 cases, the accuracy was 73.1%(49/67). (3) Therapy evaluation: According to the PRETEXT staging and CT features, 7 cases of Stage I and 24/27 of Stage II were underwent the operation directly; As 3/27 of Stage II had ascites or portal vein tumor thrombus, with 31 cases of Stage III and 2 case of Stage IV, were underwent tumor biopsy. 31 cases who underwent the operation directly, 7 cases were followed up and had no recurrence, 24 cases were underwent chemotherapy and had no recurrence too. 36 cases were received 2 durations of chemotherapy, all 36 tumors were reduced in size, but the PRETEXT staging had no change, then the tumors were received resection. Conclusions: PRETEXT staging system combined with MSCT is valuable for evaluaton the hepatoblastoma and conduting the timing of surgery and the chemotherapy protocols.

Ultrasound has limited diagnostic utility in children with acute lymphoblastic leukemia developing pancreatitis. Purpose or Case Report: Acute lymphoblastic leukemia (ALL) is the most common childhood malignancy. Children with ALL are at risk for developing acute pancreatitis (AP) during treatment, most commonly related to asparaginase. According to the American College of Radiology guidelines, ultrasound (US) should be the first line imaging modality in the diagnosis of AP. However, AP in children with ALL is thought to be due to direct pancreatic injury rather than ductal obstruction, and thus US may not be the optimal imaging modality for this diagnosis.

This retrospective study was approved by the institutional review board. Protocol databases were searched for ALL patients who were diagnosed with AP during therapy, according to Common Terminology Criteria for Adverse Events (CTAE) version 3. This list was crossreferenced with the diagnostic imaging database to identify patients who had undergone abdominal US or CT within 10 days of AP diagnosis. Chemotherapy dosing, amylase/lipase levels, clinical symptoms, and dates of imaging studies were recorded. All CT and US studies were overread by a radiology trainee blinded to the original imaging report, for findings of AP according to the CT Severity Index (CTSI) and the Revised Atlanta Classification. Discrepancies in the diagnosis of AP were adjudicated by a pediatric radiologist. Results: 69 patients, ranging from 2-21 years, experienced 88 episodes of AP (between [2008] [2009] [2010] [2011] [2012] [2013] [2014] [2015] [2016] [2017] [2018] and underwent 98 US and 44 CT exams. 72/88 (82%) events occurred within 30 days of asparaginase administration. 69 events (69/88, 78%) were diagnosed clinically by the presence of abdominal pain and amylase/lipase levels greater than 3 times the upper limit of normal. No imaging was obtained in 18 (20%) of events. The pancreas was completely obscured in 12/98 (12%) of US exams and was never visualized in entirety by US. The overall sensitivity for the detection of AP was 47% by US. Although obtained less frequently, CT detected AP in all but one case (98% sensitivity). CTAE Grade 4 events had the highest CTSI Purpose or Case Report: To analyze the use of implanted porta-caths in Children with Medical Complexity (CMC), neurological impairment and difficult venous access. Methods & Materials: REB approved retrospective single centre observational study of port-a-caths placed by Interventional Radiology (IR), in CMC with neurological impairment, to meet their general vascular access needs. Details of peripheral intravenous (PIV) attempts, PIV successful starts, PIV complications, alternative devices, port-a-cath insertions, dwell times, times accessed, and port-a-cath-related complications were analyzed. Information for the year pre-porta-cath was compared to the year post-port-a-cath in 21 patients (10M; 11F). The Wilcoxon signed ranks test were performed by SPSS. Results: 21 patients underwent 26 port-a-cath insertions (median age 3 years; median weight 12.7 kg). The median porta-cath dwell time was 26.5 months, total number of port-a-cath dwell days was 31,632 and total number of days accessed was 1,066 (median 21/pt). There was a significant reduction (p<0.05) in number of PIV attempts, PIV starts, needle pokes, ED visits and admissions in year the post-port-a-cath compared to pre-port-a-cath. There was no severe adverse events, 6 moderate and 18 mild adverse events. Limitations of the study include its retrospective design, lack of scores for difficult IV access or scales for IV infiltration and extravasation. Conclusions: Port-a-cath placement in CMC significantly reduced the number of PIV attempts, needle pokes and admission requirements with no major added complications. These results provide useful information for parents and health care providers considering port-a-cath placement in this population. Purpose or Case Report: Stenosis of the inferior vena cava or the hepatic veins is an underlying cause for hepatic compromise in liver transplant patients and in patients with inflammatory diseases of the liver, leading to considerable morbidity. CTguided or ultrasound-guided minimally invasive vessel recanalization procedures can reverse liver failure in these patients by restoring vessel patency. However, it can be difficult to employ these techniques using traditional guidewire tips against fibrotic obstructions associated with vessels that have undergone repeated stenosis. Sharp recanalization offers an alternative strategy to gain access through these obstacles by utilizing needles or the sharp end of the guidewire tip to penetrate the vessel obstruction. This technique has been welldescribed and employed successfully in adult patients, but has not yet been reported in the pediatric setting. Results: Here, we present 3 cases of image-guided hepatic vein sharp recanalization in pediatric patients, one of whom required stent placement for recurrent venous stenotic obstructions of the inferior vena cava and hepatic veins. We were able to restore hepatic vein patency in all three patients with no acute or longterm complications. Two of the patients had undergone previous liver transplants, and in following the durability of the venous recanalization in these patients, we observed long-term extension of liver graft life to the extent of at least 18 months and 38 months, respectively.

To the best of our knowledge, we present the first application of sharp recanalization in pediatric patients, and demonstrate evidence that it can be used safely and effectively to restore IVC and HV patency and prolong liver function in this patient population. Purpose or Case Report: Percutaneous CT-guided biopsy and ablation are established techniques to diagnose and treat solid tumors. However, some tumors are not adequately visualized with CT, even after IV contrast administration, because of transient enhancement. Many of these "CT-occult" tumors are optimally seen on preprocedural MRI or PET scans. To enable targeting of such tumors, we developed software that fuses MRI/PET images to intraprocedural CT in real time. The software corrects for soft-tissue deformations which occur due to differences in patient positioning or motion. The fusion imaging is generated after each intraprocedural CT scan and takes advantage of all available imaging data. We report retrospective evaluation of this fusion method for pediatric interventional radiology procedures.

The software was tested on four patients with CT-occult tumors undergoing biopsy. Two patients had bone lesions (Salmonella osteomyelitis and Ewing Sarcoma) that were optimally visualized on preprocedural MRI and PET. Two patients with neurofibromatosis had soft-tissue lesions with regions that were suspicious for malignant transformation best seen on preprocedural MRI and PET. During biopsy, CT images were pushed from the scanner to the workstation and fused with the corresponding MRI or PET. The time and accuracy of image fusion were measured. Accuracy was measured as target registration error (TRE) at anatomic landmarks identified by an expert in individual images while blinded to the fusion images. Results: Clinical experts reviewed the fusion images after the procedures and judged them to be visually adequate and accurate. The mean TRE of 4.1 mm supported this finding. The TRE data further showed that the algorithm achieved subvoxel registration accuracy (registration error < voxel size) of 0.85 voxels, averaged over all four cases. The mean time of image data transfer from CT scanner to the fusion workstation was 2.0 s for volumetric and 0.4 s for fluoroscopic scans. The mean time to perform image fusion was 35.7 s for volumetric scans and 7.0 s for fluoroscopic scans. Conclusions: The image fusion software was fast and allowed accurate visualization of MRI/PET visible tumors on the intraprocedural CT images. When fully integrated in the IR workflow, this tool has the potential to provide improve targeting of CT-occult tumors and improve efficiency and efficacy for CT-guided biopsy and ablation. A prospective clinical trial is planned to quantify the anticipated clinical benefits.Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: There is limited understanding and utilization of dynamic contrast enhancement MRI of marrow, periosteum and cartilage for diagnosis of musculoskeletal (MSK) disease in children. This is partly due to limited availability of pediatric disease models of marrow inflammation, infection, infiltration, or involvement by tumor. Herein we explore signal intensity time curves of relevant MSK targets using a 3D Golden-angle RAdial Sparse Parallel (GRASP) MRI technique. GRASP is an accelerated, freebreathing dynamic acquisition that has been shown to reduce the need for sedation. We aim to establish a baseline for normal enhancement characteristics of marrow, cartilage, synovium and periosteum of the growing skeleton in a sheep model as a precursor to translation to children. Methods & Materials: GRASP data were acquired in 4 sheep using standard dose gadobutrol at 3 mL/sec. All studies were performed on a 3T system. GRASP data are acquired with consecutive radial spokes that are rotated by the golden angle. The data can be reconstructed into time-resolved dynamic frames with user-selectable temporal resolution during contrast passage (https://cai2r.net/research/radial-vibe-sequence). We reconstructed at a temporal resolution of 4.5s. Signal intensity curves were generated from regions-of-interests in: marrow of diaphysis, metaphysis, and epiphysis of the humerus, proximal humeral physis, periosteum of the proximal humerus, a thoracic vertebral body and an adjacent intervertebral disc. Results: A sample data set is here https://www.dropbox.com/s/zmh5dakn847s6w0/20180727.mov All structures exhibited onset of contrast-enhancement within approximately the same time frame after contrast administration, with a time to peak signal of <10s and with little washout over 180 seconds. There was no difference in the slope of wash-in and wash-out. However, the peak signal varied by anatomy. Specifically, signal in the diaphyseal marrow increased from 16% to 31% above baseline; metaphyseal marrow: 56% to 246%; physis: 155% to 206%; epiphyseal marrow: 54% to 96%; periosteum: 62% to 339%; intervertebral disc: 65% to 118%; and vertebral body: 155% to 321%. Conclusions: GRASP allows rapid free-breathing characterization of contrast ehancement in the growing skeleton. It provides excellent anatomic delineation and can potentially demonstrate difference in peak signal between relevant MSK targets. We hope to translate our technique to children, and utilize it for qualitative and quantitative diagnosis of pediatric MSK disease. Purpose or Case Report: To date, a standard for CRD has not been established. Plain radiographs are a cost-effective, low-risk option for studying osseous structures in great detail. It is well known that bone age is superior to chronological age for determining biological and structural maturity; for decades, pediatricians have relied on plain films of the wrist and hand for assessing bone age in children (e.g. Greulich & Pyle Atlas, Tanner Whitehouse Method). However, some have proposed new approaches with greater accuracy and reliability across all ages and ethnic groups. Here, we present a new standardized radiographic dimension, CRD, as a tool to assess developmental maturity, which can be readily obtained from standard medical imaging platforms. Methods & Materials: X-rays of the elbow were systematically searched in the electronic medical record at a single institution from a 15-month period (June 1, 2017 -Sept 30, 2018). Inclusion criteria included patients <18 years of age with plain films in anteroposterior (AP) and lateral views. Patients with radial head dislocation, displaced supracondylar fractures, or a history of dysplasia were excluded. Follow up images were also excluded. CRD values, measured as closest linear distance from the humeral capitellum to the radial head, were measured by a staff radiologist. P-values <0.5 using a two-tailed Student's ttest were deemed to be statisticallysignificant. Results: 75 patients ages 3 months to 18 years were included in the study. 45 were male (mean age=10 yrs, range=24 mo-18 yrs) and 29 were female (mean age=9 yrs, range=3 mo-17 yrs).CRD decreased with age, with the most significant drop off observed after age 5, when most patients developed a radial head ossification center. Mean AP CRD for ages 5 and below was 7.25 mm [SD ±1.54 mm], while mean AP CRD for ages 6 and above was 3.41 mm [SD ±0.81 mm] (P <0.0001). Conclusions: The standard for determining developing skeletal maturity has historically involved X-rays of the hand and wrist. However, these methods have shown inconsistency across certain age groups and ethnic groups. Here, we present another radiographic measurement to aid in the determination of skeletal maturity. CRD proves to have a strong negative curvilinear relationship with increasing age, particularly from ages 3 months to 5 years, before many healthy patients have developed a radial head ossification center and thus demonstrate largest CRD values. Knowledge of this measurement will aid diagnosis of radial head subluxation. (n=1)] were reviewed retrospectively. Results: Most common presenting symptoms were soft tissue swelling and pain (n=13, 65%) while fever was accompanied only in three patients (15%). Median duration of symptom was 1 month (range: 2 days-3 months). Leukocytosis and elevation of C-reactive protein (CRP, > 0.5 mg/dL) were depicted in three (15%) and seven patients (35%), respectively. Sixteen children (80%) received Tokyo-172 vaccine by percutaneous multiple puncture method, three (15%) and one (5%) received intradermal Tokyo-172 and Danish strain, respectively. Distal femur (n=7) was the most frequently involved site followed by tarsal bones (n=4), proximal tibia (n=2), distal humerus (n=1), distal radius (n=1), sternum (n=1), rib (n=1) and proximal phalanx of big toe (n=1). Epiphyseal involvement of the long tubular bones were seen in seven patients (35%). Multiple bone lesions were depicted in two patients. Plain radiography showed ill-defined osteolytic lesion with cortical disruption in 13 patients. MRI showed central necrosis in the involved marrow cavity with cortical destruction and cold abscess formation in 14 patients (70 %). Surgical drainage was performed in 19 patients (95%), and half of them required repeated surgical interventions due to recurrent infection. Antituberculosis medications were administered for a median duration of 12 months (range, 12-31 months). Conclusions: BCG osteitis in immunocompetent children is a rare but serious complication of BCG immunization. High level of suspicion of BCG osteitis based on clinical and radiological manifestations is important for early diagnosis and prompt management. Purpose or Case Report: Children with Duchene muscular dystrophy (DMD) have increased risk for osteoporosis and fragility fractures due to progressive muscle weakness with eventual loss of ambulation and chronic corticosteroid therapy. Their fracture prevalence is 40-60%, and increases with age. Bone Mineral Density (BMD) is not reflective of fracture risk. Recent studies showed that cortical thickness and area were associated with increased facture risk. Digital X-ray measurement of the cortical thickness of the metacarpal bones can be used as a marker for metacarpal cortical health. To our knowledge, this marker of bone health has not been evaluated in DMD patients. The purpose of this study was to assess bone age (BA), cortical thickness and area of metacarpal bones in relation to age in children with DMD.

In this retrospective IRB approved study, 47 boys (age range: 3-17 years; mean:11.9 ± 4.5 years) with DMD who had a bone age radiograph was included. Automated determination of bone age, and cortical bone density expressed as Bone Health Index (BHI) were performed using BoneXpert (Visiana, Holte, Denmark). Validation of bone age from Bonexpert has been performed extensively before. Zscores were computed for bone age and BHI, and age-related changes were noted. Pearson's correlation coefficient (r) was calculated for each metric in comparison to age. Results: The mean z scores of bone age and BHI reduced significantly with age. Z-scores for BA and BHI were: 1) 5-8 years (n=7), BA:-0.51; BHI: -1.17; 2) 8-11 years (n=14), BA: -1.17; BHI: -1.78; 3) 11-15 years ( n=21), BA: -2.36; BHI: -1.745; 4) 15-17 years (n=4), BA:-3.77, BHI:-2.46. The Pearson's r = -0.47 for BA and r = -0.36 for BHI (reduces with age, p<0.005). The actual BHI with age was: 1) 5-8 years BHI: 3.86 ± 0.38; 2) 8-11 years; BHI: 3.83 ± 0.35; 3) 11-15 years BHI: 3.76 ± 0.95; 4) 15-17 years ; BHI: 3.84 ± 0.32 Conclusions: We showcase a new metric that looks at bone health in DMD children, and demonstrate a worsening BHI with age in DMD patients. BHI may have values in assessment of bone health in children with DMD. The worsening BHI with increasing age may correlate with increased fracture risk that is known to increase over time due to disease progression and longer corticosteroid exposure. Future study is needed to determine the association of BHI and fracture risk. Purpose or Case Report: We describe a process for preoperative virtual surgical planning and creation of patient specific surgical guides for bony tumor resection in pediatric orthopedic surgery and demonstrate a case in which this process was used for successful surgical guidance.

Herein we describe the general process for creating a virtual surgical plan (VSP) and show an example of a scapula tumor resection aided by a 3D printed surgical guide.First, a virtual 3D model of the bony anatomy is reconstructed using pre-operative CT scan images.MRI sequences that clearly delineate the tumor margins then are chosen to reconstruct the tumor.Using segmentation, region growing and 3D modeling algorithms of MIMICS TM (Materialise,Belgium), 3D models of the region of Purpose or Case Report: To determine if a 3-minute Goldenangle RAdial Sparse Parallel (GRASP) dynamic contrast enhanced (DCE) MRI sequence with 8-10 second temporal resolution is equivalent in diagnostic performance to a conventional MR sequences (CS) in the assessment of osteomyelitis (OM) in children.Methods & Materials: 6 subjects (9m -23y) with suspected OM were imaged at 3T as part of an IRB-approved study. The scans included both conventional sequences (T1, IR, PD, T2weighted and T1 fat suppressed postcontrast) for osteomyelitis as well as a 3-minute GRASP DCE sequence that we have been concurrently evaluating in a sheep model. CS were interpreted per clinical standards. GRASP was interpreted independently by 3 radiologists, blinded to CS, and scored with binary (y/n) answers for reporting elements in OM including marrow signal abnormality, synovitis, subperiosteal abscess, intraosseous abscess, soft tissue or muscle hypoenhancement, myositis, and cellulitis. Results: Average total exam time for CS protocol was 46.2 ± 15.1 minutes. Average additional time for GRASP was 8 ± 5.1 minutes including prescription and scanning. Average slice thickness was 3.2 mm for CS and 1.5 mm for GRASP. CS demonstrated marrow signal abnormality in 4 of 6 cases, attributed to OM in 3, with one case being posttraumatic. Synovitis (n=3), soft tissue hypoenhancement (n=3), myositis (n=2), and cellulitis (n=4) were also identified. There were no cases of intraosseous or subperiosteal abscess. Based on the average sensitivity and specificity of the 3 readers, GRASP was relatively sensitive (83%) and highly specific (100%) in detection of marrow signal abnormality, highly sensitive (100%) and specific (89%) for soft tissue or muscle hypoenhancement, and 100% sensitive and 67% specific for cellulitis. GRASP was less sensitive than CS for detection of synovitis (50%) and myositis (50%), with high specificity (100% and 92% respectively). Locations of disease were concordant across readers and consistent with CS. Conclusions: Based on preliminary data in this ongoing study, GRASP holds promise for replacing CS in the diagnosis of OM and ultimately reducing sedation. Further analysis is necessary to determine if the addition of a single precontrast fluid sensitive sequence and quantitative DCE curves may provide additional diagnostic benefit.

Haris Majeed, BSc 1 , Marshall Sussman, PhD 2 , Brian Feldman 1 , Carina Man 1 , carina.man@sickkids.ca; Victor Blanchette 1 , Andrea Doria, MD 1 ; 1 Diagnostic Imaging, The Hospital for Sick Children, Toronto, Ontario, Canada, 2 Toronto General Hospital, Toronto, Ontario, Canada Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: To evaluate the T2 map relaxation times in healthy male children and adolescent ankle cartilage under altered magnetic resonance image (MRI) protocols and to document trends in these T2 relaxation times with varying ages and body mass index (BMI).

This cross sectional study recruited 11 healthy male children and adolescents (median age of 14 years; range 8-17 years), who each underwent 3.0 Tesla T2 map MRI examinations of ankle (tibia-talus) using three protocols. Protocol 1 -a single echo spin echo constant TR/TE (TE = 13, 19, 28 ms; TR = 513, 519, 528 ms). Protocol 2 -a multi-echo spin echo (TE = 9.6, 19.2, 28.8 ms; TR = 1000 ms). Protocol 3 (high spatial resolution) -a multi-echo spin echo (TE = 11.1, 22.2, 33.3 ms; TR = 1690 ms). Images were analyzed at sagittal lateral and medial MRI slices using a house-made MATLAB software. A P value less than 0.05 was considered statistically significant. Results: We found statistically significant negative associations between age and T2 relaxation times of ankle cartilage, ranging from -0.91 (P < 0.001) to -0.66 (P = 0.03) for all image protocols. Furthermore, statistically significant negative trends of T2 relaxation times for ankle cartilage were found with increasing age, ranging from -2.08 ms/year (P = 0.006) to -0.80 ms/year (P < 0.01) for all image protocols. Similarly, mean T2 relaxation times were found upon using a constant TR/TE to other image protocols, thus enforcing the generalizability of the protocol. In contrast, weak associations were found between BMI and ankle cartilage T2 relaxation times. Conclusions: Age plays an important role in understanding cartilage T2 relaxation times. Additionally a TR/TE protocol can be used to help optimize scanning time for children and adolescents.

The Role of MR Susceptibility-Weighted Imaging in Acute Pediatric Seizures: in Relation to Electroencephalographic Activities Yongwoo Kim, kyw47914@gmail.com; Jae-Yeon Hwang; Radiology , Pusan National University Children's Hospital, Yangsan, Korea (the Republic of) Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: To evaluate the relationship between cortical venous signal and electrographic activity in children with seizure using magnetic resonance imaging with susceptibility weighted imaging (SWI) and electroencephalogrphy (EEG).

Children presented with seizures, who underwent both SWI and EEG within 12 hours after seizure onset were retrospectively reviewed. An increased signal of cortical veins (SWI+) was assessed using SWI, while abnormal activities such as slowing or epileptiform discharges (EEG+) were investigated on EEG. We defined three groups of patients in accordance to the topographic correlation between SWI+ and EEG+: (A) no increased venous flow and no abnormal discharges, (B) discordant finding between the SWI+ and EEG+ area, (C) concordant distribution between the SWI+ and EEG+ area. Results: We identified 297 children (194 in group-A, 76 in group-B, and 27 in group-C). The mean age among the three groups was similar (group-A =3.5±4.5 years; group-B = 3.1±3.5 years; group-C = 5.2±5.4 years, p=0.079). Multiple seizures were revealed more frequently in group-C (51.9%) than in group-A (23.7%) or group-B (39.5%, p=0.002). The incidence of newly-diagnosed epilepsy was significantly higher in group-C (20/27, 74.1%, p=0.001) than in group-A (45/194, 23.2%) or group-B (17/76, 22.4%). By contrast, there were no significant Purpose or Case Report: To evaluate diagnostic performance of quantitative gray-scale ultrasound as an objective method in evaluation of pediatric thyroiditis.

In this retrospective study of 37 children with tissue proven diagnosis, two radiologists independently reviewed thyroid ultrasounds twice and subjectively classified images according to presence or absence of thyroiditis. A consensus session was performed for patients for which there was disagreement. Unweighted kappa coefficients were calculated to assess intra-and inter-observer reliability. Pearson chi-square and Fisher's exact tests were used to compare categorical measures by final pathology.A third radiologist performed quantitative measurements of echointensity level of the thyroid and adjacent strap muscles from US images using a PACS-based tool. Thyroid /muscle ratio (TMR) was obtained by dividing thyroid mean by muscle mean values. Heterogeneity index (HI) for thyroid was calculated by dividing thyroid SD by thyroid mean values. Wilcoxon Rank Sum tests were used to assess distribution of continuous risk factors by final pathology. Results: Patient group comprised 29 females and 8 males, with median age 17 years (interquartile range 15-18). By pathology, 19 (51.3%) patients had lymphocytic thyroiditis and 18 (48%) had normal thyroid. For subjective assessment, there was fair inter-observer agreement (kappa .36 (95% CI .14-.57), p=.004) and slight intra-observer agreement for each radiologist (kappa .13 and .17, p>.05). A large proportion of patients for whom consensus review indicated thyroiditis were confirmed with pathology (12/19 (63%), p=.03). For quantitative assessment, no significant difference between thyroiditis and normal thyroid groups was found for either TMR (1.51 and 1.62, respectively, p=.82) or HI (.23 and .23, respectively, p=.37). Conclusions: Quantitative gray-scale ultrasound did not accurately diagnose thyroiditis. However, subjective consensus evaluation showed significant correlation with the condition, suggesting that multiple radiologists performing more than one review may be beneficial for accurate diagnosis of thyroiditis in children.

Matthew O. Thompson, MD 1 , matthewthomp@gmail.com; Joseph Davis, MD 1 , Atalie C. Thompson, MD 1 , Nathan Hull, MD 2 , Gary Schooler, MD 1 ; 1 Pediatric Radiology, Duke University, Durham, NC, 2 May Clinic, Rochester, MN Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: The purpose of this study is to determine whether a decrease in hematocrit is predictive of an intracranial hemorrhage (ICH) on neonatal head ultrasound (HUS).

The indication for the HUS was categorized as being related to hematocrit (e.g. "drop in hematocrit") or unrelated to hematocrit. The medical record was also reviewed for the change in hematocrit in the 48 hours prior to HUS, a decrease in hematocrit, gestational age, number of days since birth, prematurity, and abnormal neurologic exam. The association between these variables and the presence or absence of hemorrhage on the HUS was analyzed using STATA 12.1. Results: Twenty-one percent (N=29/136) of neonates had an ICH on their HUS. Studies that were ordered with an indication related to hematocrit were less likely to have an ICH on HUS compared to those with alternative indications for the study (odds ratio 0.35, p=0.018). The mean change in hematocrit (p=0.95), a decrease in hematocrit (p=0.30), and an abnormal neurological examination (p=0.25) were not associated with ICH on HUS. However, a lower gestational age (mean difference 4.47, p<0.001) and number of days since birth (mean difference 25.1, p=0.01) were both significantly predictive of ICH. Those with an ICH were also three times more likely to be premature, though this association was only borderline significant (odds ratio 3.03. p=0.09).Conclusions: Diffusion weighted imaging had no impact on distant staging in our pediatric cohort, but may add value in local staging. Routine performance of whole body DWI can be eliminated and limited to the region of interest in pediatric patients undergoing PET/MRI to improve patient throughput.

Ultrasensitive detection of translocations in the cell free DNA of pediatric sarcoma patients Heike E. Daldrup-Link, MD, PhD 1 , Avanthi Shah, MD 2 , Frederick M. Wittber, MD 1 , fwittber@stanford.edu; Tej Azad 1 , Jake Chabon 1 , Stan Leung 2 , Aviv Spillinger 2 , Heng-Yi Liu 2 , Marcus Breese 2 , Maxamilian Diehn 1 , Ash Alizadeh 1 , Alejandro Sweet-Cordero 1 ; 1 Radiology, Stanford, Palo Alto, CA, 2 UCSF, San Francisco, CA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Currently, biopsy serves as the gold standard to accurately diagnose disease in pediatric sarcoma patients, but the risks of anesthesia and surgery, along with a failure to characterize the true heterogeneity of disease make this method less than ideal. Progression and response to therapy are monitored by radiologic exams which lack the sensitivity to detect early relapse. Cell free DNA (cfDNA) is released into the plasma as they undergo apoptosis and necrosis. ctDNA represents a small fraction of cfDNA in cancer patients and contains tumor specific alterations. It holds promise as a highly sensitive and specific biomarker. A limitation in applying liquid biopsy in clinical practice is the need to develop PCR or other DNA analysis methods to detect alterations specific to a single patient. We have developed a more widely applicable off the shelf test that does not involve a patient specific design.

Our CAPP Seq technique involves designing a selector comprised of custom designed oligonucleotide probes that tile across genomic regions of interest. These oligonucleotide probes are used to enrich for the relevant ctDNA via hybrid capture, followed by ultra deep sequencing to analyze alterations in the selected regions. Our selector was applied to pretreatment plasma samples from newly diagnosed or newly relapsed pediatric sarcoma patients. Plasma samples were analyzed at key timepoints over the course of treatment. Results: Pediatric sarcoma patients had higher levels of cfDNA when compared to published levels in adult cancer patients. Canonical translocations were detected in the plasma of 13/14 (93%) pediatric sarcoma patients. This was confirmed by analysis of matched tumor samples, when available. Patients with metastatic disease had higher ctDNA levels compared to nonmetastatic patients. ctDNA levels correlated with clinicalcourse and, in some cases, rising ctDNA levels predicted relapse, earlier than was clincally apparent by imaging studies. Conclusions: ctDNA analysis holds promise as an ultrasensitive and specific tool for monitoring tumor burden. Our assay was able to detect ctDNA in the plasma of metastatic and nonmetastatic pediatric sarcoma patients at diagnosis. Furthermore, we demonstrated that ctDNA levels correlated with clinical response to therapy. In some cases, ctDNA levels proved more sensitive than imaging, detecting minimal residual disease andpredicting relapse. Purpose or Case Report: Experience in pediatric nuclear medicine is limited and not uniform across residency and pediatric fellowship training programs. At our tertiary care pediatric hospital, emergent nuclear medicine exams performed after hours are currently read by radiologists with expertise in pediatric nuclear medicine. A decision to shift this responsibility to in-house faculty has prompted creation of a continuing medical education (CME) learning module as part of an institutional comprehensive learning management system (LMS). The goal of this module is to train faculty, fellows and residents in the indications, protocols, diagnostic criteria, potential pitfalls and problem-solving techniques when reading emergent pediatric nuclear medicine exams. The purpose of this study is to better understand the volume of nuclear medicine cases on-call and the potential need for a dedicated pediatric nuclear medicine curriculum. Methods & Materials: All nuclear medicine gastrointestinal bleeding scans, Meckel's scans, hepatobiliary scans for biliary leak or acute cholecystitis, brain death scans and renal transplant evaluations performed between July 1, 2017 and June 30, 2018 were reviewed. Exams ordered after 4:30 PM on weekdays or performed on weekends were considered call studies. Faculty were surveyed to assess length of time since reading each type of scan, as well as their level of comfort protocolling and interpreting studies. Results: Case review revealed 54 emergent pediatric nuclear medicine studies with 13 (24.1%) read after hours. Of all studies performed, 28.6% of gastrointestinal bleeding studies (n= 2/7), 16.7% of Meckel's scans (n= 5/30), 22.2% of brain death studies (n = 2/9), 33.3% renal transplant scan (n=1/3), 75.0% of acute cholecystitis scans (n=3/4) and 0% of biliary leak scans (n=0/1), were read on-call. 19 (70.4%) of 27 faculty members who take call completed the survey. At least 63.2% have not read any of these types of nuclear medicine study since residency/fellowship or not at all. The percentage of those uncomfortable protocoling studies ranges from 73.7% to 89.5%, greatest with renal scans. The percentage of those uncomfortable interpreting studies ranges from 47.4% to 84.2%, greatest with renal scans. Conclusions: Even in large academic centers, there is a limited number of radiologists with expertise in pediatric nuclear medicine, making call coverage challenging. There is a need for continued training to provide important emergent nuclear medicine studies for patients after hours. Purpose or Case Report: To evaluate the diagnostic performance of 3 Tesla lung magnetic resonance imaging (MRI) in children with allergic bronchopulmonary aspergillosis (ABPA).

This study protocol was approved by the institutional ethics committee and written informed consent was obtained from parents/ guardians. From October 2015 to January 2018, we prospectively evaluated twenty-seven consecutive children with ABPA. The diagnosis of ABPA was made on the ISHAM-ABPA working group criteria. High resolution computed tomography (HRCT) and 3T MRI of the chest was performed on the same day. Bronchiectasis, consolidation, nodules, and mucus impaction were assessed in all segments. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of MRI were calculated using HRCT findings as the reference standard. Interobserver agreement was calculated using the kappa statistic. Results: The mean age of the patients was 9.89 years (range: 5-16 years). There were 20 males and 7 females. The sensitivity, specificity, PPV, and NPV for bronchiectasis was 68%, 100%, 100% and 71.43% respectively. The sensitivity, specificity, PPV, and NPV for consolidation was 80%, 100%, 100% and 96% respectively. For detection of nodules, the sensitivity, specificity, PPV, and NPV was 75%, 100%, 100% and 88.46% respectively. There was 100% sensitivity, specificity, PPV and NPV for mucus impaction. There was a high degree of interobserver agreement for MRI findings (k=0.9-1) as well as agreement (k=0.7-1) between CT and MRI for all the four findings. Conclusions: With the currently available routine MR sequences, MRI demonstrates high specificity but less sensitivity and negative predictive value to HRCT scan in children with ABPA. Newer MR sequences need to be explored and validated to enhance the potential of lung MRI in ABPA.

Diagnostic accuracy of ultrasound for identifying metastatic cervical adenopathy in pediatric patients with differentiated thyroid carcinoma at presentation Mara Navallas Irujo, MD, maria.navallasirujo@sickkids.ca; Alan Daneman, Afsaneh Amirabadi, Jonathan Wasserman; Hospital for Sick Children, Toronto, Ontario, Canada Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: To evaluate the diagnostic accuracy of US and the most useful sonographic features for diagnosing metastatic cervical adenopathy in pediatric patients with thyroid carcinoma Methods & Materials: IRB-approved retrospective study in a tertiary children's hospital. Eligiblity for inclusion were all consecutive children with pathologically proven thyroid carcinoma from 2008 to 2017. Patients with no preoperative US or no resected lymph nodes were excluded. Pathology report of lymphadenopathy was used as the gold standard. Size, shape, echotexture and vascularity of the lymph nodes were analyzed and compared to the pathology findings Results: We reviewed preoperative US and histology reports of resected lymph nodes in 52 children with proven thyroid carcinoma (33F,19M; age range 5-18y; mean 13y). Metastatic cervical lymph node disease was documented on US in 29 children and on histology in 33. Sensitivity of US was 79%, specificity 84%, PPV 90%, NPV 70% and accuracy 81%. A significant association was seen between round shape, echotexture, vascularity and lymph node histology. The measurements in short axis for the metastatic nodes was significantly higher than benign nodes (U=168.5,P=0.005, Mann-Whitney U test). No significant difference was noted between the groups in long axis. Logistic regression Univariate analysis showed that round shape (OR=0.054,95%CI=0.006-0.466), echotexture (OR=0.048,95%CI=0.01-0.234), vascularity (OR=0.025,95%CI=0.003-0.225) and short axis measurement (OR=1.473,95%CI=1.019-2.130) contribute significantly to make a positive diagnosis. Multiple logistic regression analysis showed only vascularity contributed significantly to explain the disease probability when adjusting for the other variables. Importantly, 11 patients (38% of children with metastatic disease documented on US) had US diagnosis of abnormal lymph nodes based solely on abnormal echogenicity and vascularity, with normal size and shape Conclusions: Neck US showed high accuracy, sensitivity and specificity for identifying malignant adenopathy in children with thyroid carcinoma. Most of the abnormal lymph nodes were round in shape and had abnormal echogenicity and vascularity. However, this paper emphasizes that metastatic nodes may be normal in size and shape and the abnormality may be based solely on abnormal echogenicity and vascularity. Size, particularly long axis measurement, is not a reliable criterion to differentiate between malignant and benign lymph nodes. This has not been reported previously in pediatrics Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity. Purpose or Case Report: To prospectively investigate the diagnostic accuracy and added value of lung MRI for evaluating pulmonary hydatid disease in children by comparing MRI findings with MDCT findings. Methods & Materials: 28 consecutive children with clinically suspected of having pulmonary hydatid disease were enrolled in this prospective research study from October, 2012 to July, 2018. In all 28 pediatric patients (24 boys, 4 girls; mean age, 8.93 +/-3.1 years; range, 5 to 17) were included in this study, MRI without contrast and enhanced MDCT of the chest were performed within 48 hours of each other. Two pediatric radiologists independently evaluated the lungs for the presence 3 mm, corroborating the ultrasound findings and confirming the diagnosis of SMA syndrome. The patient was initiated on nutritional support and at most recent follow-up was responding well to conservative treatments. Conclusions: Knowledge of the imaging findings of this rare disorder can provide early diagnostic capabilities and lead to more effective treatment plans. Purpose or Case Report: Gadolinium, a type of MRI contrast is typically used when imaging pediatric cardiac patients. Gadolinium enhances the images to give more clarity to the radiologists when distinguishing normal tissue from abnormal tissue. Further, contrast also allows for better visualization of the great vessels.Gadolinium alters the magnetic properties of nearby water molecules in the tissue. While contrast itself does not alter the tissue composition, it does affect how the scanner takes the picture. Accordingly, images can differ between preand post-contrast administration. While IV contrast can be administered safely, it is not without logistic implications and/or risks.MRI contrast can be given to pediatric cardiac patients intravenously, but IV access is oftentimes challenging in pediatric cardiac patients. In an effort to reduce the number of unsuccessful IV attempts and quantity of contrast administered, a novel non-contrast MR sequences is being utilized to better depict the great vessels in these children. If IV access is unsucessful or the patient cannot complete the MRI, this modality is employed to obtain more information on the patient's cardiact vasculature.This purpose of this case report paper is to show the MR imaging when using a non-contrast sequence to identify cardiac vessels in pediatric patients who cannot recieve IV contrast.

Teela M. Durfee, AAS, teela13@hotmail.com; Tara Cielma; Ultrasound, Children's National Medical Center, Silver Spring, MD Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Pediatric scrotal ultrasound pathology can be difficult because of the similar presentation of different pathology such as scrotal torsion, epididymitis, inguinal hernia, epididymal appendix torsion, and trauma.Identifying key factors of each condition is paramount to providing a clear diagnostic picture in the setting of an atypical presentation of any pathology.Epididymitis is an inflammatory process precipitated by bacterial or viral infections. Symptoms typically present as increasing scrotal pain with swelling of the epididymis. Depending on degree of severity imaging presentation can mimic other pathology.A 5 year old male transferred to a pediatric facility from an area hospital to assess suspected incarcerated hernia with outside CT and ultrasound images.

Outside diagnostic imaging shows complex, hyperemic mass-like structure in the superior aspect of the right scrotal sac with surrounding edema and preserved flow to both testicles.Outside CT exam, upon second read, shows an enlarged, edematous right epididymis and small hydrocele with overlying soft tissue thickening and stranding tissue extending from the right scrotum through the inguinal canal. Examination shows no evidence of hernia or incarceration. Results: Additional diagnostic sonography imaging demonstrated progression of inflammatory condition within the right scrotal and inguinal area. The right epididymis presented edematous, hyperemic, and heterogenous with surrounding soft tissue edema and scrotal wall thickening. Flow to the right testical was preserved however, the inflammation affected the orientation of the right testical in the scrotal sac. Conclusions: This case highlights the difficulty in clearly identifying pathology for acute testicular pain in a pediatric patient without known trauma. Purpose or Case Report: Imaging requisitions are the legal documents that health professionals use to communicate to radiologists and Medical Radiation Technologists (MRT(R)s) the relevant clinical information to guide their requested radiographic examination. Inadequate or incomplete information may have a substantial impact on patient care.Since the implementation of our electronic medical record system, EPIC (Epic Systems Corp.), we have the ability to track the number of errors on requisitions; the most common error is having the incorrect body parts selected or multiple unnecessary exams being ordered. The MRT(R)'s identify these errors and correct the inconsistencies following departmental protocols. Without such vigilance on the MRT(R)'s part, inappropriate radiographs could have been performed and could lead to additional, unnecessary radiation exposure. To help select the most appropriate protocol, educate and guide the health professionals at our hospital, an electronic software tool, "EDI" has been created.EDI (Examine the patient, Determine the correct radiograph, Input the order with pertinent and relevant clinical information) is an interactive tool that includes the exam protocol with the associated views per body part. Each exam also includes EDI with the field of view that will be exposed to radiation during the exam.Our goal is that EDI will serve health professionals to better understand what order needs to be selected in the electronic system and which radiographs will be provided. Ultimately, this will reduce any unnecessary exams and reduce pediatric radiation exposure. Purpose or Case Report: Purpose: The long-term effects from gadolinium tissue deposition in organs and the brain is unknown especially in children who may need repeated contrast studies for necessary follow-up of cardiovascular abnormalities. We aim to show the utility of a non-contrast MRA sequence in older children and young adults undergoing chest and/or abdominal MRA. Methods & Materials: Methods: We retrospectively review the utility of non-contrast navigator-triggered and ECG-gated 3D steady-state free precession (SSFP) imaging of the chest and abdomen generally acquired in the coronal plane performed for evaluation of cardiovascular abnormalities. Results: Results: When acquired during diastole to limit the amount of turbulent flow, non-contrast 3D SSFP imaging is successful in all patients who are able to remain still with regular breathing rates. The studies are diagnostic in over 90% of all older children and young adults with non-diagnostic imaging only occurring when there is substantial flow turbulence related to vascular or valve stenosis. The pulmonary veins are also occasionally not well visualized. There can be substantial artifact in younger children. A contrast-enhanced study should be performed following a non-diagnostic noncontrast study which will occur in a small percentage of older children and young adult patients. Conclusions: Conclusions: Gadolinium based contrast can be eliminated for most chest and abdomen MRA studies in older children and young adults. 

Bowel ultrasound studies at our institute were reviewed retrospectively with selected representative cases chosen to illustrate technical aspects and clinical indications for this procedure. Correlation was made with follow up radiology studies, clinical and/or surgical outcomes.Necrotizing enterocolitis, inflammatory bowel disease, infectious colitis, neutropenic enterocolitis, malrotation, short bowel syndrome, appendicitis, intussusception, Meckel's diverticulitis, intestinal polyp, vascular malformations, graft versus host disease, intestinal hemorrhage, perforation, and obstruction will be discussed. Results: Typical findings in normal and abnormal pediatric bowel exams will be illustrated. Emphasis on appropriate technique and methodology for serial examinations will be described.Conclusions: Through this exhibit, participants will learn to utilize bowel ultrasound as a powerful tool in evaluating pediatric gastrointestinal diseases and understand how to optimize their technique. Purpose or Case Report: Ultrasound has become a powerful tool for the evaluation of bowel pathology in the children. In the newborn, ultrasound (US) can be used to diagnose, stage, and follow up necrotizing enterocolitis (NEC) and its complication. US has excellent sensitivity and specificity for the identification of bowel wall thickening, peristalsis, pneumatosis, portal venous gas and free air and has become an integral evaluation tool for the newborn with abdominal distention. Our educational poster will:1. Summarize proper US equipment, technique and protocol for diagnosing and following up NEC2. Discuss advantages and benefits of incorporating US in the evaluation of bowel in newborns3. Illustrate the sonographic findings of mild, moderate and severe NEC through cases Methods & Materials: Pictorial review of US findings of NEC including grayscale and color Doppler. Comparison is made to similar findings in radiographs and clinical correlation. Results: The spectrum of normal and abnormal neonatal bowel US findings are presented. Abnormal findings are presented using illustrative cases. Conclusions: Ultrasound is a powerful tool that supplements and at times can replaced more traditional diagnostic methods in the assessment of NEC. This educational exhibit provides a practical guide for systematic evaluation of bowel in the newborn and how to diagnose, stage and follow up NEC and its complications. Purpose or Case Report: In 2016, we focused on MRI safety and the inherent cultural barriers to reduce the risk of undetected or misplaced metal objects causing MRI accidents. We successfully addressed this opportunity using a coordinated approach with Patient Safety, Performance Improvement and Radiology Leadership to provide a multi-faceted solution.Despite implementation of ferromagnetic detection systems (FMDS) technology, numerous gaps in screening effectiveness were identified. Three primary improvement objectives were established involving place, people and process leading to 42 new practice changes that were implemented, and 68 existing process improvements instituted. Alarm fatigue was one among many identified risks. Variables included the physical location of the projectile on the transport person, as well as the horizontal or vertical orientation of potential hazard while being carried were identified during a series of nine standardized PDSA testing sessions that were completed in the clinical setting.

Histiocytosis. In order to confirm the importance of follow up skeletal surveys, I will present cases and associated radiographs that show signs of LCH prior to treatment and post treatment.

Parker T. Stanley, MHA, BSRST 1 , parker.stanley.us@gmail.com; Erika Rubesova, MD 2 ; 1 Virginia Commonwealth University Health, Charlottesville, VA, 2 Lucile Packard Children's Hospital, Palo Alto, CA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Introduction: Juvenile Idiopathic Arthritis (JIA) is the most common rheumatic entity in children and the second most common cause of musculoskeletal symptoms in pediatric patients. Treatment of JIA largely depends on frequently managing inflammation within the joints, and as such, ultrasound is a prime imaging modality that can be used for detection and monitoring of inflammation. Ultrasound is more sensitive than plain films in the early detection of JIA, and ultrasound is more sensitive than clinical examinations alone. As a low-cost, high-resolution imaging modality, ultrasound allows visualization of the joint spaces, can be used to detect joint effusions, and can monitor synovial proliferation and joint hyperemia, all while not exposing the patient to radiation. We will present ultrasound images of various joints in children such as knees, ankles, wrists, hands and feet. Ultrasound technique, imaging planes, choice of probes frequencies and ultrasound settings are analyzed for optimization of the images. This presentation will highlight the diagnostic imaging findings of JIA on ultrasound, such the presence of joint effusions, synovial proliferation, capsular thickening, tendinitis as well as provide information on common pitfalls associated with age-specific appearances of pediatric musculoskeletal structures on ultrasound. Familiarity with JIA and musculoskeletal ultrasound findings will allow timely diagnosis and implementation of appropriate treatment strategies. Purpose or Case Report: The MAGEC System is a surgical treatment for children with severe spinal deformities. The system utilizes surgically implanted rods, which are periodically extended by an External Remote Control (ERC). The traditional method to measure lengthening progression involved radiation exposure. Ultrasound provides an immediate evaluation with no bio-effects. Additionally, soft tissue changes incurred by rod implantation can be seen, which is difficult to assess with a radiograph. The sonographer acquires both pre-and postlengthening measurements before and after the ERC is applied to the child's skin at the level of the rods. The entire process can be completed in minutes within a single outpatient room, for the ease of both patient and parent. Methods & Materials: Children with scoliosis too severe for bracing or casting and too young for spinal fusion, have been traditionally managed with growing rods. This involves repeated surgical elongation, often requiring ten or more surgeries before the child reaches skeletal maturity.MAGEC rods require initial implantation surgery and are expanded with the use of an externally applied ERC. The sonographer scans each rod, measuring the pre-lengthening measurement. The physician utilizes the ERC to externally manipulate the MAGEC rods. The sonographer scans the child again, and can determine if the rods have been sufficiently extended. Additional ERC applications, if necessary, can then be employed. Results: Ultrasound has proven to be a useful tool in the measurement of MAGEC rod extension in patients with severe spinal deformities. It can easily be performed within the examination room in conjunction with the application of the ERC. Conclusions: The use of ultrasound to measure the lengthening of MAGEC rods minimizes radiation exposure, decreases the need for repeated surgeries, and allows for soft tissue evaluation. It streamlines the process of obtaining measurements and can be utilized in one room within an outpatient setting, simplifying the encounter for patient and family.

Marcy L. Hutchinson, AS 1 , hutchinsonm@email.chop.edu; Brandi Kozak, BSDI 1 , Victor Ho-Fung, MD 1 , Nancy Chauvin, MD 2 ; 1 Children's Hospital of Philadelphia, Philadelphia, PA, 2 Penn State Hersey Radiology, Hersey, PA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Background:Sonographic techniques are simple and safe. The utility and clinical value of dynamic sonographic evaluation of the glenohumeral joint in infants with brachial plexus injury (BPBI) has been described since the late 1990's. However, this technique remains an underutilized tool in most imaging practices. Brachial plexus injury during the birthing process can lead to glenoid dysplasia, posterior shoulder subluxation and significant morbidity if left untreated. Imaging evaluation of the degree of deformity is paramount to guide clinical treatment and the follow-up of complicated cases. Imaging techniques include MRI, CT and US. Shoulder ultrasound provides a dynamic, noninvasive method of evaluation. It has also been our experience that this imaging technique can be mastered by the Sonographer and supported with accurate interpretation by the Radiologist.Purpose:Describe the imaging features of glenohumeral joint dysplasia due to BPBI with emphasis on ultrasound technique.Review step-by-step dynamic sonographic evaluation of the glenohumeral joint in BPBI with emphasis on correct positioning of the patient to ensure consistency and reproducible quality imaging.Improve both confidence and independent evaluation of patients through this presentation by the sonographer.

Proper technique will be demonstrated utilizing multiplanar dynamic imaging techniques performed with high frequency linear transducers. Tips for proper patient positioning in the abducted and adducted shoulder positions will be included. The normal glenohumeral joint anatomy will be shown as well as cases of varying degrees of shoulder dysplasia. Purpose or Case Report: Medical imaging of the soft tissue of the neck or airway is one of the most common, and sometimes challenging, exams performed at a pediatric hospital. Imaging of the airway can help identify enlarged adenoids, the presence of a foreign body, an abscess, pathology or anatomical abnormality. It is crucial to understand the proper technique, breathing instructions and positioning in order to obtain optimal imaging for correct diagnosis. Different radiographic findings will be presented with examples of incorrect positioning, grid usage, exposure factors, artifacts and ways to improve the image quality. This poster will also provide technologists with tips and recommendations to help them to feel prepared and confident to work with all pediatric patients. Purpose or Case Report: The goal of this presentation is to provide a review of the important role Dual Energy X-ray Absorptiometry (DXA) plays in treating pediatric patients with low bone density. It also aims to educate on the different DXA machines used today, give a review of the official positions for pediatric DXA scans determined by the International Society for Clinical Densitometry (ISCD), give a brief history of the lateral distal femur (LDF) scan and why it was developed, and touch on the other modalities that can be used to asses bone density..Osteoporosis is commonly thought of as an adult bone health issue; however recently it has gained more attention as being an issue for pediatric patients. Different factors affect bone health, such as genetics and family history, diet and exercise, certain medications, and whether a patient is ambulatory or non-ambulatory. It is important to conduct the proper testing to determine a patient's fracture risk and trying to prevent further bone deterioration. A DXA Scan is the gold standard across all age groups in providing vital bone health information as it provides a look into the body with minimal radiation exposure to evaluate the bones to determine how much at risk a patient may be.

We will look at examples of properly positioned scans and what the results can look like. We will also look at the difference between ambulatory and non-ambulatory lateral distal femurs scans and compare the Z-Scores of those scans.

Transcranial Doppler -How to successfully perform an optimal TCD on our younger patient demographic (ages 2-5 years)Brandi Kozak, BSDI,kozakb@email.chop.edu; Ultrasound, Children's Hospital of Philadelphia, Philadelphia, PA Disclosures: All authors have disclosed no financial interests, arrangements or affiliations in the context of this activity.Purpose or Case Report: Transcranial Doppler examinations are crucial in the care and treatment of patients in the Sickle cell population. The earlier we are able to obtain diagnostic studies on these patients the more effective treatment is and can reduce the risk of the patient suffering from a life altering stroke.

This procedure while not invasive can be very frightening for a younger patient. We employ several different techniques and distractions tools to help the patient cope soundly through the exam. We work closely with the Hematology department staff and have developed a TCD task force specifically focused for this age group to increase our success rate. Some of the strategies we employ are as follows: arranging the ultrasound room so that the patient is able to view the television to watch a favorite show or program, letting the child play games on smart phones or tablets (without too much patient movement), having a parent or guardian lying with or holding the patient, sitting the patient up slightly, blowing bubbles, singing, having hand held musical or light up toys available and/or arranging to have a child life specialist present for the duration of the exam to help distract the patient. We have also produced a short informational video for new patients to view prior to coming to their first TCD exam so they can see what to expect. Results: By having several options in place and available, by planning ahead and having resilient, dedicated sonographers we are able to successfully perform TCDs on younger and younger patients thereby facilitating more positive outcomes for this patient population. Conclusions: Transcranial Doppler ultrasound is a low cost, easy to perform with dedicated practice, portable, radiation-free modality that with set parameters in place can be extremely successful in imaging younger patients. This exam gives immediate, important information to the Hematologist that will improve long term outcomes in the Sickle cell population.

Paper 156