key: cord-0984219-n9ugmg4f
authors: Ma, Depeng; Gao, Ronghui; Li, Minghui; Qiu, Jianfeng
title: Mechanical and medical imaging properties of 3D‐printed materials as tissue equivalent materials
date: 2021-12-08
journal: J Appl Clin Med Phys
DOI: 10.1002/acm2.13495
sha: 96f6908b6d5ae874980e660c1433098485526533
doc_id: 984219
cord_uid: n9ugmg4f

Three materials of polylactic acid (PLA), polyamide 12 (PA12), and light curing resin (LCR) were used to construct phantom using 3D printing technology. The mechanical and medical imaging properties of the three materials, such as elastic modulus, density, effective atomic number, X‐ray attenuation coefficient, computed tomography (CT) number, and acoustic properties, were investigated. The results showed that the elastic modulus for PLA was 1.98 × 10(3) MPa, for PA12 was 848 MPa, for LCR was 1.18×10(3) MPa, and that of three materials was close to some bones. In the range of 40∼120 kV, the X‐ray attenuation coefficient of three materials decreased with increasing tube voltage. The CT number for PLA, PA12, and LCR was 144, −88, and 312 Hounsfield units at 120 kV tube voltage, respectively. The density and the effective atomic number product (ρ*Z (eff)) were computed from three materials and decreased in the order of LCR, PLA, and PA12. The acoustic properties of materials were also studied. The speeds of sound of three materials were similar with those of some soft tissues.

Tissue equivalent materials have been widely used in medical research and clinical simulator to mimic the properties of real tissues. For instance, medical imaging researchers utilize tissue equivalent materials to calibrate equipment and develop new imaging methods. 1, 2 In clinical simulators, tissue equivalent materials play important roles as idealized tissue model to train clinical skills of medical worker. 3, 4 In the above mentioned, tissue equivalent materials were generally constructed into the phantom. The manufacturing method of phan-phantoms is high in costs and is limited to accommodate personalized patient's pathological features. In addition, such phantoms consist of homogenous materials do not simulate the inhomogeneity of different tissue such as bone, muscle, and lung. 3 3D printing, also called additive manufacturing, is a process that produces objects by adding material in layers. This layer-by-layer production method provides greater flexibility and creativity in the design process. 3D printing can significantly speeds up the design and prototyping process. Therefore, 3D printing-related medical devices have been rapidly applied in many medical fields. The surgical simulation system for anatomical disease models is one of the most promising areas for the clinical application of 3D printing technology. 8, 9 The patient-specific disease model which is manufactured using 3D printing technology can ensure the precise definition of the disease scope and the detailed reflection of the relevant anatomical structure. [10] [11] [12] [13] Many studies on the clinical application of 3D printing technology have been reported, such as models cranioplasty 14 and cerebral aneurysm models 15 in neurosurgery, patient-specific instruments within diagnosis and treatment, [16] [17] [18] [19] and nasopharyngeal swabs for diagnosis of COVID-19. [20] [21] [22] For meeting the requirements of clinical application,the mechanical and medical imaging properties of 3D printing materials must be close to those of real tissue. Unfortunately, compared with conventional tissue equivalent materials such as agarose, 23 PVC, 24 and gelatin, 25 the reports for properties of 3D printing materials, especially medical imaging properties, were still relatively rare. In this paper, the mechanical and medical imaging properties of three 3D printed materials of polylactic acid (PLA), polyamide 12 (PA12), and light curing resin (LCR) were systematically tested and analyzed. These results would make us more familiar with 3D printing materials, as well as providing clear guidance in selecting materials suitable as fat, bone, and soft properties.In addition,these data could also be used as a reference tool to guide the multi-modal imaging of 3D-printed materials.

Three kinds of 3D printing materials of PLA, PA12, and LCR, were purchased from Creality, China. All five colors of PLA were 1.75-mm filament. The five colors were yellow, red, black, fluorescent red, and fluorescent green. PA12 and LCR were white powder and milky liquid, respectively.

PLA samples were fabricated using a fused deposition modeling 3D printer (uPrint SE, Stratasys, Israel). Selective laser sintering 3D printer (S360, UnionTech, China) and stereolithography (SLA) 3D printer (HT600S, Hontai, China) were employed to process PA 12 powder and LCR resin, respectively.

The 3D printing material samples with 30 × 10 mm 2 and a length of 30 mm were made for X-ray attenuation coefficient (μ) testing. The method to measure μ of the sample was through digital radiography (CXDI-55G, Canon, Japan) with dosimeter (Solid dose 400, German). 26 The μ was determined using the following equation:

where x is the thickness of the sample, I 0 and I are Xray beam intensity before and after transmitted through a sample of thickness of x. A CT scan (Lightspeed, GE, USA) was used to measure CT number of 3D printing material samples at the same X-ray tube voltage. The elastic modulus and density of the samples were measured according to ASTM D638 and D792 at 25 ± 2 • C. The speed of sound and acoustic attenuation coefficient of the samples were measured using 5 MHz ultrasound transducer (AFG3102, Tektronix, USA) in water tank following GB/T15261-2008.

The imaging properties of X-ray imaging and ultrasound imaging of three 3D printing materials were measured. The X-ray attenuation coefficient (μ) of three materials was tested and calculated for tube voltage from 40-120 kVp at the 100 mA tube current. The μ of different materials decreased from 46.8 to 15.6 m -1 with increasing tube voltage (Figure 1 ). At the same time, it could be seen from Figure 1 that the μ of LCR was the biggest among three kinds of 3D printing materials, followed by PLA and lastly PA12. The effect of color on μ has also been studied in the same material, and results showed that less change of μ values was observed when the color of PLA changed ( Figure 2 ). The CT number is expressed in terms of Hounsfield units (HU) corresponding to the X-ray attenuation. CT numbers are computed as follows:

where μ m is the measured linear attenuation coefficient of the material, μ w is the attenuation of water, and K (1000) is the scaling factor. Hence, materials that attenuate more than water have positive CT numbers, whereas materials with less attenuation than water have negative CT numbers. CT numbers of materials at 120 kV ranged from −88∼144 HU. CT number for PA12 was −88 HU, PLA was 144 HU, and LCR 312 HU. The speed of sound and acoustic attenuation coefficient of PLA, LCR, and PA were measured ( Table 6 ). The speed of sound of PLA, about 2246 m/s, is close to that of PA12 (about 2242 m/s) and slightly lower than LCR (about 2427 m/s). However, significantly differences in acoustic attenuation coefficients of above mentioned materials were observed. The acoustic attenuation coefficient for PA12 was 5.44 dB/cm/Hz, LCR was 3.55 dB/cm/Hz, and PLA was 2.31 dB/cm/Hz. For compound or polymer, their atomic number is called the effective atomic number (Z eff ). The Z eff could be estimated by CT measurements using two different X-ray tube voltages. 27 Therefore, the Z eff of 3D printed materials was computed using the method described by Noblet et al. 28 Eleven tissue substitute materials of known densities and elemental compositions were scanned with CT to construct the stoichiometric relationship of the function of ρ and Z eff versus CT number. The curve ρZ eff versus CT number led to a monotonic relationship (Equation 3).

where H is CT number of material. The CT numbers of three materials were put into Equation 1, and The ρ*Z eff product of the materials was obtained,for LCR 9.85,PLA 8.84, and PA12 7.41. Compared with traditionally manufactured, the stressstrain curves of three 3D printed materials exhibited no yield point and were linearly. Elastic modulus is the ratio of stress to strain. As shown in Table 5 , elastic modulus of materials ranged from 848∼1980 MPa. The elastic modulus for PA12 was 848 MPa, LCR was 1180 MPa, and PLA 1980 MPa.

The μ energy dependency curves for material were also plotted against kVp and shown in Figure 1 . As shown in Figure 1 , the μ of all materials decreased with increasing tube voltage. Photoelectric absorption and Compton scattering contribute substantially to X-ray attenuation in the energy range used in diagnostic radiology when Xray transmitted through substance. Moreover, these two interactions decrease with the increase of X-ray energy. Another phenomenon observed was μ(LCR) > μ(PLA) > μ(PA12) in the 40-120kVp range at 100 mA. In addition to tube voltage and tube current, the properties of the material (such as effective atomic number Z eff , density ρ) also affect X-ray attenuation. 29 In general, substance with higher Z eff or ρ will absorb more X-rays, resulting in more X-ray attenuation. Therefore, we inferred that the Z eff and ρ of three materials might have order of Z(LCR) > Z(PLA) > Z(PA12), and/or ρ(LCR) > ρ(PLA) > ρ(PA12).

The μ change values of PLA with different colors are provided in Figure 2 . Less change in the μ values was observed when color of PLA changed. Specific material formulations are typically not provided by manufacturers, and, discussing formulations was also outside the scope of this work. However, we could still conclude that TA B L E 1 The CT numbers of polylactic acid (PLA), polyamide (PA), and light curing resin (LCR) at 120 kV and 100 mA 

The CT number is a normalized quantifier of tissue density which is imaged in CT scanner. CT number is obtained by scaling substance μ value to more convenient integers and normalizing to μ value containing water. Furthermore, CT numbers are displayed as grayscale pixels on the digital image. Clinically, the CT number may be of more relevance than μ for characterizing the materials. The CT numbers of PLA, PA12, and LCR were measured and are listed in Table 1 . The data in Table 1 showed that the CT number of LCR is the largest, followed by PLA, lastly PA12 among the three materials. The CT numbers of LCR and PLA were positive, indicating that X-ray attenuation of two materials was stronger than that of water. PA12 had a negative CT number and weaker X-ray attenuation than water. Table 2 presented CT numbers of some soft and skeletal tissue. The CT number of LCR was 312 HU compared with spongiosa at 262 HU and sternum at 385 HU. PLA and PA12 had value of 144 and −88 HU, comparable with cartilage (102 HU) and adipose tissue (−77 HU),respectively. These suggested that 3D-printed materials could construct medical phantoms to be equivalent to real tissues.

Several factors affect X-ray attenuation. Some are related to the X-ray beam and the others to properties of the material through which the radiation is passing. The properties of material include the thickness, TA B L E 3 The ρ and Z eff of polylactic acid (PLA), polyamide (PA), and light curing resin (LCR) the density (ρ), and the atomic number (Z) of material. Table 3 lists the ρ and Z eff of three materials calculated following Equation 1. As depicted in Table 3 , the density descending order of materials was PLA, LCR, PA12; and Z eff was the descending order of LCR, PLA, and PA12. The density or atomic number of the material decreases, the attenuation produced by a given thickness decreases. The ρ*Z eff product was also computed from three materials (Table 3 ) and decreased in the order of LCR, PLA, and PA12. The order of ρZ eff agreed with the results obtained from μ measurements. The ρ and Z eff of body tissues from International Commission on Radiation Units and Measurements report 46 are listed in Table 4 . 31 The ρZ eff of PLA was close to human muscle, skin, and heart. The ρZ eff of PA12 was comparable with red marrow and gastrointestinal (GI) tract. The ρZ eff of LCR was similar with cartilage and thyroid.

The stress-strain curves of three 3D-printed materials are presented in Figure 3 . As shown in Figure 3 , the stress of three materials increased with the strain linearly. Stress-strain curves for materials did not exhibit yield point, and the elongation at break was very low (about 0.05%). These effects might be due to layerby-layer production method was prone to separation between layers during the tensile test. To get the elastic modulus, E, the measured stress data with strain in the linear elastic region of the material were used. Table 5 lists the elastic modulus of three materials and five types of tissues. PLA and PA12 had the largest and smallest elastic modulus among the three materials, respectively. In addition, all E values were seen to be above 100 MPa. Most normal soft tissues (liver, muscle, etc.) have elastic modulus on the order of 10 kPa. 32 Bone has elastic modulus, more than 10 MPa. Therefore, three 3D-printed materials were not within the scope of soft tissues. But, it was satisfactory that elastic modulus of three materials was close to bone tissue.

The acoustic properties (speed of sound, c, acoustic attenuation coefficient, α) of three samples are shown in Table 6 . Three kinds of materials had similar speeds of sound. However, there were obvious differences in acoustic attenuation coefficient of materials. Most tissues in human have speeds of sound around 1500-1600 m/s (Table 7) . Compared with human tissues, c of samples was close to those of human skin. However, the acoustic attenuation coefficients in samples were between 2.21 and 5.44 dB/cm/Hz, which had a gap with that of real tissue.

Mechanical and medical imaging properties of three 3Dprinted materials (PLA, PA12, and LCR) were measured. Compared with human tissues, the elastic modulus of the three materials was relatively high (>100 MPa) and could be used to simulate bones. X-ray attenuation coefficient of materials decreased with increasing tube voltage. And, the order of μ(LCR) > μ(PLA) > μ(PA12) in the 40-120kVp range was observed. The difference of color would not cause a significant change in μ value. For CT number, PLA and LCR were comparable with bone tissue when PA12 was close to adipose. The order of the density and the effective atomic number product (ρ* Z eff ) agreed with the results obtained from μ measurements. The ρZ eff of LCR was similar to that of some bones; PLA and PA12 were close to some soft tissues. The acoustic properties of materials had a gap with that of real tissue. However, speeds of the sound of samples were close to those of human skin. Therefore, 3D printed materials could be used as tissue equivalent materials to simulate some soft and bone tissues, when using 3D printing technology to construct a phantom.

The authors gratefully acknowledged the financial support from the Academic Promotion Programme of Shandong First Medical University (2019QL009),the Qingdao Innovative Talents Projection, the Taishan Scholars Program of Shandong Province (TS201712065) and Science and Technology funding from Jinan (2020GXRC018).

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Writing the first draft of the manuscript: Depeng Ma. Collection and analysis of data: Depeng Ma, Ronghui Gao, and Minghui Li. Revising and final approval of the manuscript: Jianfeng Qiu.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Development of phantom materials with independently adjustable CT-and MR-contrast at 0.35, 1.5 and 3 T

Fabrication of a pediatric torso phantom with multiple tissues represented using a dual nozzle thermoplastic 3D printer

A systematic review on 3D-printed imaging and dosimetry phantoms in radiations therapy

Endourological training using 3D-printed bladder phantoms: development and prospective evaluation

Molecular interactions and forces of adhesion between single human neural stem cells and gelatin methacrylate hydrogels of varying stiffness

Gellan gum-based gels with tunable relaxation properties for MRI phantoms

Agarose-based tissue mimicking optical phantoms for diffuse reflectance spectroscopy

Clinical application of patient-specific 3D printing brain tumor model production system for neurosurgery

3D printing of physical organ models: recent developments and challenges

Cerebral aneurysm clipping surgery simulation using patient-specific 3D printing and silicone casting

Threedimensional printing in radiation oncology: a systematic review of the literature

Clinical applications of patient-specific 3D printed models in cardiovascular disease: current status and future directions

Identification of robust and reproducible CT-texture metrics using a customized 3D-printed texture phantom

On-demand intraoperative 3-dimensional printing of custom cranioplastic prostheses

Comparison of two three-dimensional printed models of complex intracranial aneurysms for surgical simulation

Office 3D-printing in paediatric orthopaedics: the orthopaedic surgeon's guide

3D-printed headrest for frameless Gamma Knife radiosurgery: design and validation

Three-dimensional printing of medical devices used directly to treat patients: a systematic review. 3D Print Addit Manuf

Ninjaflex vs superflab: a comparison of dosimetric properties, conformity to the skin surface, planning target volume coverage and positional reproducibility for external beam radiotherapy

Prospective clinical validation of 3D printed nasopharyngeal swabs for diagnosis of COVID-19

Performance of threedimensional printed nasopharyngeal swabs for COVID-19 testing

Pandemic printing: a novel 3D-printed swab for detecting SARS-CoV-2

Design and fabrication of a personalized anthropomorphic phantom using 3D printing and tissue equivalent materials

Characterizing mechanical and medical imaging properties of polyvinyl chloride-based tissue-mimicking materials

Developing and characterizing MR/CT-visible materials used in QA phantoms for MRgRT systems

Fabrication of an anthropomorphic heterogeneous mouse phantom for multimodality medical imaging

Measurement of effective atomic numbers using energy-resolved computed tomography

A new tissue segmentation method to calculate 3D dose in small animal radiation therapy

Gamma and X-ray shielding properties of various types of steels

Correlation between CT numbers and tissue parameters needed for Monte Carlo simulations of clinical dose distributions

Report 46. Photon, Electron, Proton and Neutron Interaction Data for Body Tissues. International Commission on Radiation Units and Measurements

Soft tissue elastometer

Basics of Biomedical Ultrasound for Engineers