CCID-27902-emerging-technologies-for-the-detection-of-melanoma---achiev


© 2012 Herman, publisher and licensee Dove Medical Press Ltd. This is an Open Access article  
which permits unrestricted noncommercial use, provided the original work is properly cited.

Clinical, Cosmetic and Investigational Dermatology 2012:5 195–212

Clinical, Cosmetic and Investigational Dermatology

Emerging technologies for the detection  
of melanoma: achieving better outcomes

Cila Herman
Department of Mechanical 
Engineering, Johns Hopkins University, 
Baltimore, MD, USA

Correspondence: Cila Herman 
Department of Mechanical Engineering, 
Johns Hopkins University, 102 Latrobe 
Hall, 3400 North Charles Street, 
Baltimore, MD 21218-2682, USA 
Tel +1 410 516 4467 
Fax +1 410 516 7254 
Email cherman@jhu.edu

Abstract: Every year around 2.5–3 million skin lesions are biopsied in the US, and a fraction 
of these – between 50,000 and 100,000 – are diagnosed as melanoma. Diagnostic instruments 

that allow early detection of melanoma are the key to improving survival rates and reducing 

the number of unnecessary biopsies, the associated morbidity, and the costs of care. Advances 

in technology over the past 2 decades have enabled the development of new, sophisticated 

test methods, which are currently undergoing laboratory and small-scale clinical testing. This 

review highlights and compares some of the emerging technologies that hold the promise of 

melanoma diagnosis at an early stage of the disease. The needs for detection at different levels 

(patient, primary care, specialized care) are discussed, and three broad classes of instruments are 

identified that are capable of satisfying these needs. Technical and clinical requirements on the 

diagnostic instruments are introduced to aid the comparison and evaluation of new technologies. 

White- and polarized-light imaging, spatial and spectroscopic multispectral methods, quantitative 

thermographic imaging, confocal microscopy, Optical Coherence Tomography (OCT), and 

Terahertz (THZ) imaging methods are highlighted in light of the criteria identified in the review. 

Based on the properties, possibilities, and limitations of individual methods, those best suited for 

a particular setting are identified. Challenges faced in development and wide-scale application 

of novel technologies are addressed.

Keywords: Infrared imaging, thermography, melanoma detection and diagnosis, quantitative 
imaging, in vivo diagnostics

Introduction and background
Melanoma incidence is increasing at one of the fastest rates for all cancers in the US, 

with a current lifetime risk of one in 55.1 Every year, approximately 68,000 melanomas 

will be diagnosed, with around 8700 resulting in death. If melanoma is detected at an 

early stage, before the tumor has penetrated the epidermis, the 5-year survival rate 

is about 99%.2 However, the 5-year survival rate drops dramatically – to 15% – for 

patients with advanced disease.2,3 At present, there are no systemic treatments available 

that significantly extend the life span of patients with advanced melanoma;4,5 therefore, 

the key to extended survival is early detection and treatment.6,7 In order to enable early 

detection and diagnosis, avoid unnecessary biopsies, and ultimately reduce the cost of 

care, it is essential to develop accurate, sensitive, and objective quantitative diagnostic 

instruments that have the potential to have a deep impact on the disease. Dramatic 

advances in solid-state electronics and instrumentation, imaging hardware, computers, 

and software tools, including image-analysis techniques, have opened new avenues 

for quantitative imaging for the detection of melanoma. This review highlights and 

Dovepress

submit your manuscript | www.dovepress.com

Dovepress 
195

R E v I E w

open access to scientific and medical research

Open Access Full Text Article

http://dx.doi.org/10.2147/CCID.S27902

 
C

lin
ic

a
l, 

C
o

sm
e

tic
 a

n
d

 I
n

ve
st

ig
a

tio
n

a
l D

e
rm

a
to

lo
g

y 
d

o
w

n
lo

a
d

e
d

 f
ro

m
 h

tt
p

s:
//

w
w

w
.d

o
ve

p
re

ss
.c

o
m

/ 
b

y 
1

2
8

.1
8

2
.8

1
.3

4
 o

n
 0

6
-A

p
r-

2
0

2
1

F
o

r 
p

e
rs

o
n

a
l u

se
 o

n
ly

.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

mailto:cherman@jhu.edu
www.dovepress.com
www.dovepress.com
www.dovepress.com
http://dx.doi.org/10.2147/CCID.S27902


Clinical, Cosmetic and Investigational Dermatology 2012:5

compares some of the emerging technologies that hold the 

promise of melanoma diagnosis. The emphasis of the review 

is placed on the technology aspect of the instrumentation, 

recent advances, challenges, and research needs.

While most cancers develop deeper in the body and cannot 

be viewed with the naked eye, melanoma most often develops 

on the skin and is easily visible. The challenge that remains 

is recognizing it early, so that treatment can be initiated 

before the tumor reaches a thickness of 0.76 mm.8 Detecting 

melanoma with the human eye can be challenging, because 

of the diversity of possible presentations during the early 

stage of the disease. While trained dermatologists can reliably 

recognize advanced melanoma by visual inspection (relying 

on the Asymmetry, Border irregularity, Color variegation, 

Diameter . 6 mm, Evolution [ABCDE] criteria) melanomas 

at their very early and most curable stages often mimic 

histologically benign look-alike lesions. Instruments that can 

capture and quantify information not readily available to the 

human eye and aid in discerning whether a lesion is benign 

or malignant could dramatically improve survival rates and 

reduce health-care costs. Early detection of melanoma relies 

on periodic (yearly) complete and thorough examinations by 

physicians, supplemented by more frequent (monthly) self-

exams by patients. Both physicians and patients would benefit 

from accurate and reliable instruments able to yield objective 

and quantitative assessment of lesions. These scanners can 

serve as preliminary screening tools, allowing more time for 

the clinician and specialist to focus on lesions that are more 

likely to be malignant.

Types of melanoma detectors based  
on end user
Therefore, the long-term goals of research efforts when devel-

oping instruments for the noninvasive in vivo diagnosis of mela-

noma are to develop three key types of devices (scanners):

•	 Type I: Relatively low-cost, easy-to-use, handheld scanners 
for screening purposes (for self-exams of at-risk patients or 

screening in primary care facilities), providing an indication 

whether a lesion has a potential for malignancy (in this 

case, the patient would be advised to seek medical care 

with advanced diagnostics methods). Currently, full-body 

photography is the tool available for patients’ self-exams, 

and this method relies on the subjective visual comparison 

of the real-life lesion and the previously photographed 

lesion. Modern technologies offer a range of more objective 

methods to aid the self-exams of patients.

•	 Type II: Precision scanners that provide accurate and 
detailed visual and/or quantitative characterization of 

individual lesions for use in specialized clinics by trained 

professionals, to aid decisions regarding the need for biopsy 

or the extent of surgery. The focus of current research efforts 

is the development of these instruments, and several promis-

ing techniques have emerged over the past two decades. The 

transition from laboratory demonstrations and small-sample 

pilot trials to clinical practice remains a challenge.

•	 Type III: Full-body-imaging systems that would ideally map 
the entire body surface of the patient in three dimensions, 

with sufficient resolution to provide three-dimensional 

(3D) information similar to the currently used digital 

full-body photography. This imaging system would serve 

as a catalog of all lesions. White-light imaging could be 

combined and coupled with one of the more advanced 

and sensitive imaging systems with advanced diagnostic 

capabilities. These imaging systems are essential to 

provide objective reference baseline information when 

tracking the evolution of lesions; however, they also hold 

the promise of diagnostic capabilities.

With advances in telecommunication and information 

technologies, telemedicine, the technology of providing 

clinical health care at a distance is becoming a reality. 

All three types of scanners would support telemedicine 

applications by allowing the development of an image 

repository for a patient as well as the tracking of the patient’s 

lesions over time and enable long-term follow-up care. 

Since melanoma is becoming a widespread health problem, 

telemedicine combined with quality information from 

scanners would be an invaluable tool in geographic regions 

where specialized care is not easily accessible. It would also 

help in eliminating distance barriers and improving access 

to medical services that would often not be consistently 

available.

Requirements
Ideally, a melanoma-detection instrument (scanner) should 

satisfy a range of requirements:

1.  The scanner should provide an objective assessment of 

the lesion by measuring some physical properties of the 

tissue (by quantitatively comparing healthy and  diseased 

tissue) and lesion, or by comparing the parameters of the 

diseased tissue with a previously established scale in 

order to eliminate the subjective component in the 

 interpretation of the results.

2.  The outcome of the scanning and data-analysis process 

should be delivered in quantitative form (for example, 

by characterizing the malignant potential of skin lesions 

with a simple, numeric, quantitative scale), followed by 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

196

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

a conclusion/recommendation regarding the nature of 

the lesion.

 3.  In its basic form (type I and type III scanners used for 

screening), the screening scanner would not require 

a trained radiologist for interpretation of the images. 

Data and images from a type II device would be both 

quantitative and qualitative, and allow analysis of 

the characteristic features of a lesion in more detail. 

Specialized training would be necessary for the 

interpretation of data delivered by type II scanners.

 4.  The device and the measurement method should be 

sensitive enough to deliver a strong signal even for very 

early stage melanoma lesions.

 5.  The measurement method should be robust and insen-

sitive to differences in ambient and measurement 

conditions and local and individual variations of skin 

properties.

 6.  Ideally, measurement should be relative, in that it 

compares the properties of the lesion and the healthy 

tissue of the same subject, rather than comparing 

the responses of different individuals. In this way, 

the influence of individual variations of skin-tissue 

properties can be greatly reduced or even eliminated.

 7.  The scanner should be easy to use without sophisticated 

and specialized training (except for some applications 

of type II scanners), so that it is suitable for self-exams 

of patients between appointments.

 8.  Ideally, the device would not require perfect contact 

with the skin or the use of an immersion fluid for 

imaging.

 9.  The measurement should be noninvasive, by either not 

requiring irradiation of the tissue or keeping irradiation 

low, to minimize affecting or causing damage to the skin. 

Some scanning methods require, for example, irradiation 

with light, sound, or electromagnetic (EM) radiation, 

while others measure the radiation naturally emitted by 

the skin/human body. Other interactions with the skin 

of the patient, eg, thermostimulation by cooling, should 

be moderate, in order to avoid or minimize discomfort 

during the scanning process.

10.  Ideally, the same measurement method should be used 

for scanning of individual lesions (type I and type II 

devices), as in a full-body scanner (type III).

11.  The exclusion criteria should not limit the use of the 

device significantly.

12.  The scanner should be reliable in operation, small in 

size and be low-cost, sensitive in terms of measure-

ment capabilities, and robust in use, with a strong 

support base for manufacturing, customer support, 

and repair.

13.  The scanner(s) should be easy to integrate into 

telemedicine applications and systems.

14.	 	The scanner(s) should have the ability to interface with 
the Internet in order to deliver information and images 

into the patient’s record.

15.  The scanner should have the ability to interface with the 

manufacturer’s database and update the image library 

as new lesions are scanned. The interaction with the 

manufacturer’s system should also allow seamless 

updates of the software as the database and software 

are updated (as opposed to purchasing upgrades every 

couple of years, for example). As the calibration 

database is continuously enhanced and upgraded, the 

improvements should be made available to the end user 

continuously and seamlessly.

16.  The technology should be readily available commercially 

within a short time (a few years, depending on the length 

of the FDA approval process) and should be reliable and 

relatively low-cost.

The imaging systems currently under development as 

candidates for melanoma detection satisfy different sets of 

requirements from this list: no single technology can satisfy 

all requirements at the same time. The emphasis in this 

review is on technologies that offer the potential of in vivo 

measurements, since in vivo diagnostic tools are essential 

and very much in demand in many fields of medicine and 

in melanoma diagnosis in particular. In general, they can 

provide guidance for surgical interventions by detecting 

lesion margins or allow for replacement of general biopsies 

of suspicious tissues by more targeted biopsies (typically 

one out of 30–40 lesions biopsied is a melanoma). In this 

way, in vivo imaging methods would reduce unnecessary 

tissue excisions and the associated pathology costs, as well 

as biopsy-associated risks and morbidity. Other clinical 

applications of these imaging instruments include the 

monitoring of the effect of therapies to tailor drug or radiation 

treatments based on individual responses of the patients. 

While a considerable amount of effort has been devoted to 

the development of type II methods and devices, much less 

effort has been dedicated to type I and type III scanners. 

The existence of an accurate and sensitive screening tool 

for self-exams of patients (similar to pressure cuffs or blood 

glucose monitors) would have a dramatic impact on the 

early detection of melanoma. Some methods are well suited 

for in vitro imaging and tests, which is useful when there 

is a need for fast processing of large numbers of samples. 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

197

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

By quantifying the imaging outcomes and screening cases 

that are not likely to be malignant, the pathologist is able to 

devote more attention to the samples that indicate possible 

malignancy.

Classification of detectors based  
on detection method
Direct contact methods vs noncontact methods
The methods suitable for melanoma diagnosis can be 

classified as those that require direct contact between the 

radiation source, skin, and receiving optics, and methods 

that do not require direct contact (remote methods). 

Direct-contact methods sometimes require immersion fluids 

between the skin and the optics to avoid losses caused by 

the mismatch of the optical properties of the media through 

which the signal propagates. The use of immersion fluids 

adds complexity to the measurement, and these methods 

are typically not suitable for full-body scanners. Noncontact 

or remote measurements can be affected by the involuntary 

movement of the subject or even breathing, and they will 

require motion-tracking algorithms to be included into the 

image-analysis and processing steps when high-accuracy 

measurements are needed. Another classification is based on 

the need to apply external forcing to the investigated system 

before or during a measurement.

Active vs passive methods
Active methods involve forcing (heating, cooling, application 

of pressure, application of a tracer, or a contrast medium, 

etc) and measure the response of the system to the applied 

forcing. Passive methods characterize the system as is, ie, 

they measure steady-state properties. Passive methods are 

easier to apply; however, active methods may be faster, yield 

a stronger signal, and be more sensitive and accurate. The 

diagnostic method may require exposing the subject to an 

external source of radiation (X-rays, ultrasound, light, and 

measuring the reflected, scattered, or transmitted portion of 

that  radiation) or measurement of the radiation naturally emit-

ted by the  subject. The latter method is easier to apply and 

does not require a radiation source (making the instrument 

smaller and less complex); however, there is little flexibility 

in influencing the nature or magnitude of the measurement 

signal.

Image-forming vs quantitative methods
Diagnostic methods in medicine can be classified as image-

forming (information characterizing the lesion is contained 

in an image, such as a digital photograph, ultrasound image, 

or X-ray image) and quantitative methods that deliver a 

quantitative signal (for example, in the form of electric 

current or potential difference). Visual inspection using the 

ABCDE criteria is essentially an image-forming method, 

with the visual information regarding the characteristics of 

light reflected from the skin being detected by the human eye 

and processed by the brain. The human brain is capable of 

processing large amounts of visual information quickly, and 

after suitable training comparing this information with the 

mental database of images. The diagnosis has a subjective 

component, and the experience of the person conducting 

the analysis plays a significant role. Describing a visual 

image with numbers (quantifying it) to allow rigorous 

quantitative comparison with other images is a very complex 

process. Fully emulating the visual image analysis and 

decision-making process of the human brain with a computer 

is not possible at this stage of technical development. 

However, modern computer-based image-processing tools 

can aid the quantification process by reducing the vast amount 

of information contained in visual images to a limited amount 

of quantitative data relevant for a particular diagnosis. 

Quantitative methods deliver data that are easily quantified or 

compared (such as peaks in a radiation spectrum); however, 

they lack the subjective visual detail that visual information 

can easily convey.

An essential component of a melanoma-detection 

system is a calibration database that is proprietary to the 

manufacturer of the instrument. This database evolves as 

more information (test data) becomes available during the 

development and use of the instrument. The calibration 

database is established by comparing the measurement signal 

to the gold standard – the pathology result – for the evaluated 

lesion. The accuracy of detection is expected to improve 

over time as new lesions are added to the library of lesions. 

The lesion-classification outcomes will vary from method 

to method, and it is essential to “teach” the computer how 

to discriminate lesions based on the data accumulated in the 

calibration database. The lesion-classification outcomes are 

established using biostatistics methods. An example of this 

kind of database is the one established for MelaFind based 

on clinical studies.

Stages of instrument development
Instrumentation for the early detection of melanoma is 

complex, both in terms of hardware and software as well as 

clinical testing, and the development of a new detection or 

imaging system is a multidisciplinary effort accomplished 

in several stages:

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

198

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

1. The first step is the proof-of-concept study, involving 

the development and testing of a laboratory prototype 

instrument in a laboratory environment on an experimental 

phantom, in vitro or on animal models.

2. This step is followed by modifying the hardware for clinical 

studies and demonstrating the feasibility of the method in 

patient studies with a limited number of subjects. Initially, 

the image and data analysis are manual; after the initial 

learning phase, the objective is to develop automated 

analysis and classification criteria in the development steps 

to follow, which is a challenging task.

3. Successful feasibility testing on a limited number of 

subjects is followed by the development of equipment 

suitable for clinical testing on a larger number of patients 

to establish the calibration database. Automated image-

processing algorithms and data-processing software are 

developed based on the data collected in the calibration 

database (image library). The creation and initial 

validation of the database, software, and the decision-

making criteria is an iterative process requiring testing 

on a large number of patients and assessment of a large 

number of benign and malignant lesions.

4. Once both the software and the hardware have reached 

the necessary level of technological readiness (reaching 

the accuracy and sophistication required for clinical 

evaluation), the FDA approval process can begin. In the 

US, the first step in the approval is the clinical assessment 

of the instrument to gather data for premarket approval 

(PMA) report and documentation.

5. After receiving PMA, the market launch of the instrument 

follows. For example, it took around 10 years for MelaFind 

to mature from early proof-of-concept efforts to PMA.

There are numerous hurdles along the path of new medical 

device development, and the journey is difficult, costly and 

time-consuming. In spite of the many sophisticated laboratory 

imaging systems developed for melanoma imaging described 

in the literature, the automated quantitative detection 

remained an elusive goal, primarily because of the cost, 

effort, and time needed to develop a sophisticated, sensitive, 

and reliable device. The recent launch of the MelaFind 

scanner for the detection of melanoma is a step in the right 

direction. The development of screening instruments remains 

a challenge for the future, and it will require significant 

investment of resources, both public and private, to evaluate 

different technologies in order to develop instruments best 

suited for a particular application. It appears that justifying 

the need for investing in diagnostic tools for skin cancer, 

which is directly visible on the skin surface, is more difficult 

than for other conditions that cannot be directly visualized 

by the naked eye.

Electromagnetic radiation  
in medical diagnostics
EM radiation is the most commonly used means of probing 

human tissue in medical diagnostic and imaging instruments 

(X-rays, Magnetic Resonance Imaging (MRI), digital 

photography, etc). EM radiation is a form of energy emitted 

or absorbed by matter, which behaves like a wave when 

traveling through space with the speed c. EM radiation is 

characterized by both electric and magnetic field components, 

which oscillate in phase perpendicular to each other and 

perpendicular to the direction of the propagation of the wave 

and energy. In vacuum, EM radiation propagates at the speed 

of light - c
0
 = 2.998 × 108m/s. EM radiation has practically 

no inertia; therefore it can be used to analyze high-speed, 

unsteady processes in the human body. Its interaction with 

living tissue depends on the properties of radiation, the 

frequency ν (or wavelength λ), and the power. For radiation 
propagating in a particular medium, these two properties 

are related as λ = c/ν. EM radiation carries energy through 
space, away from the source towards the matter with which 

it interacts.

The EM spectrum is illustrated in Figure 1, in order of 

increasing frequency and decreasing wavelength. The black-

body radiation spectra for thermal radiation (left) and IR 

radiation (right) are depicted in the top portion of Figure 1. 

The longest wavelengths in the EM spectrum are called radio 

waves, followed by microwaves, IR radiation (responsible for 

thermal effects), visible light, ultraviolet radiation, X-rays, 

and gamma rays. The intermediate portion of the spectrum 

that extends from 0.1 to 100 µm is called thermal radiation, 
and it includes portions of the ultraviolet and all of the 

visible and IR spectra, as shown in Figure 1. The human 

eye is sensitive to a relatively small window of wavelengths 

of EM radiation called the visible spectrum, which is in the 

range of 0.4–0.7 µm. The nature of the interaction of EM 
radiation with biological tissue depends on the properties of 

radiation: its power and frequency (wavelength). For longer 

wavelengths and lower frequencies (radio waves, microwaves, 

IR radiation), the interaction (and potential damage) with 

cells (and other materials) is characterized by heating effects 

associated with the radiation power. For shorter wavelength, 

higher-frequency radiation (ultraviolet frequencies and 

above), such as X-rays and gamma rays, the potential damage 

to the chemical structures and bonds and living cells by EM 

radiation can be far larger than that done by simple heating. 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

199

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

incident

Emitted

Source
Detector

Scattering

Absorbed

A B

C

Muscle

Fat

Reticular demis

Papillary dermis

Epidermis

Lesion

Reflected 

Figure 2 (A) Interaction of radiation with tissue (absorption, reflection), radiation 
source, and detector; (B) digital white light, and (C) dermoscopy images of a melanoma 
lesion, capturing the reflected portion of the incident EM radiation.

1019

0.1 Å 1 Å  1 nm  10 nm  100 nm  1 µm  10 µm  100 µm  1 mm  1 cm

1 4 10
10−5

10−5

100

105

Visible
region

0.1 0.2 0.4

5800 K

1000 K

800 K

300 K

100 K

Wavelength, λ (µM)

Wavelength, λ (µM)

Wavelength

Frequency (Hz)

G
a
m

m
a
-r

a
ys

X
-r

a
ys

U
ltr

a
vi

o
le

t

V
is

ib
le

In
fr

a
 r

e
d

M
ic

ro
w

a
ve

s

R
a
d
io

, 
T

V

L
o
n
g
-w

a
ve

s

Before

Incoming
UV rays

After

50 K

2000 K

800 K

400 K

100 K

50 K

Infrared region

0.71 4 10 40 100 300 1,000

100

105

40 100 300 6001000

 10 cm  1 m  10 m  100 m  1000 m

1018 1017 1016 1015 1014 1013 1012 1011 1010 109 108 107 106

S
p

e
c

tr
a

l 
e

m
is

s
iv

e
 p

o
w

e
r,

 E
 a

,b
(W

/m
2
 µ

m
)

S
p

e
c

tr
a

l 
e

m
is

s
iv

e
 p

o
w

e
r,

 E
 a

,b
(W

/m
2
 µ

m
)

A
M

R
a

d
a

r

F
a

r 
IR

N
e

a
r 

IR

1.4 µm 15 µm

Figure 1 EM radiation spectrum shown in terms of wavelength and frequency, characteristic ranges used in medical diagnostics. 
Note: The black-body radiation spectrum for thermal and IR radiation is shown in the top portion of the image.

This can be attributed to the ability of photons to damage 

individual molecules chemically by affecting the chemical 

bonds. Whenever using EM radiation for probing human 

tissue in vivo, the frequency (wavelength) and power (and 

their combination) have to be selected in such a way to avoid 

affecting and damaging the tissue. More power can yield a 

stronger signal, but it is also more likely to cause damage; 

therefore, safety considerations will limit the magnitude of 

the probing radiation and thereby the measured signal.

A typical instrument or imaging system for medical 

diagnostic purposes consists of the source of radiation and a 

detector. The radiation emitted by the source (human body, 

sun, light bulb, laser source, etc) is reflected, absorbed, and 

scattered by human tissue, as illustrated in Figure 2A. Some 

forms of radiation, eg, X-rays, can also be transmitted through 

the human body. The human body will also emit radiation 

in the IR domain of the spectrum (not visible to the naked 

eye), and the IR emission spectrum is shown in the top-right 

portion of Figure 1. The detector captures the reflected, 

scattered, transmitted, or emitted radiation. The reflection, 

transmission, scattering, and emission properties of healthy 

and diseased tissue will usually differ, and a sensitive detector 

will image or quantify this difference.

Radiation sources and detectors have evolved greatly 

over the past few decades, which has enabled measurements 

of very small signals, some that were too small to detect 

in the past. Therefore, the technology advances may allow 

measurements today that were impossible in the recent past. 

Depending on the wavelength range used for sensing, the 

design and complexity of the radiation source and the detector 

can vary greatly, and so do the challenges and costs involved 

with building the imaging hardware. For example, sources 

and detectors for X-ray radiation instruments in medical 

diagnostics are well established and readily available, whereas 

THz sources and detectors suitable for medical applications 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

200

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

emerged on the market relatively recently. Older technologies 

are often less costly, readily available, and already have good 

distribution and support (repair) networks. This is especially 

the case when a technology is widely used in other fields, 

such as IR imaging, with an established use in night-vision 

systems. Reliability testing and cost may initially pose a 

challenge when trying to rely on a new technology. An 

essential component of most medical instruments nowadays 

is a computer that stores, analyzes, and displays data to the 

user and interfaces the Internet for information access and 

storage. Over the past few decades, computer hardware and 

software have experienced dramatic advances, allowing 

fast processing of large amounts of complex data, which is 

essential for medical diagnostic systems. Through access to 

the Internet, safe, centralized storage and sharing (as needed) 

of large amounts of data have become a reality.

Melanoma-detection methods
The majority of cutaneous melanomas appear as pigmented 

lesions of the skin. The most effective screening tool for 

detection of atypical lesions and the standard of care, 

established 27 years ago, has been visual examination: 

assessing the lesion according to the ABCDE criteria, 

introduced in 1985.9,10 The last criterion, evolution, was added 

in 2004, after recognizing the importance of the changes in 

size and shape of the lesion in the screening process.9 The 

source of radiation is white light (from natural or artificial 

sources), and the radiation reflected by the skin is detected 

by the human eye. The diagnostic accuracy of the ABCDE 

criteria has been verified in clinical practice; however, it only 

provides qualitative guidelines for melanoma identification 

and yields high rates of false positives and frequent 

false negatives. Studies suggest that almost one out of three 

melanomas could not be diagnosed using these criteria.10 

Over the past three decades, there have been improvements 

in imaging tools for melanoma detection; however, clinicians 

still primarily rely on their eyes, experience, and clinical 

judgment. Therefore, accurate, sensitive, objective, and 

quantitative instruments to aid the diagnosis of melanoma 

are still needed in clinical practice. In addition to this, 

screening tools that would alert patients during self-exams, 

coupled with the ability to store images and data online in 

a repository for long-term follow-up, would be invaluable 

in reducing the mortality associated with the disease. The 

melanoma-detection methods discussed in this review are 

summarized in Table 1 and their key features are indicated 

in the table.

Digital photography and dermoscopy
Current in vivo imaging tools commonly used by derma-

tologists are digital photography (total cutaneous imaging 

or the imaging of individual lesions) and dermoscopy. In 

digital photography, serial images recorded over time (pho-

tographic follow-up) are compared in order to find changes 

in size, shape, and color of pigmented lesions that might 

suggest malignancy. It was found that the relative change 

of a lesion is a sensitive marker for early melanoma.4,7–10,14 

However, additional studies also showed that only 35% of 

diagnosed melanomas were identified on the basis of the 

relative change.11,12

Table 1 Melanoma detection methods considered in this review and their key features

Method Type Quantitative Image- 
forming

Active/ 
passive

Contact/ 
remote

Radiation 
type

Technology 
readiness

Digital 
photography

I, II, III No Yes Passive Remote white 
light

Available

Dermoscopy II No Yes Passive Both Polarized 
light

Available

MSS I, II Yes Yes Active Contact Different 
wavelengths

PMA

MSF I, II Yes Usually 
no

Active Contact Different 
wavelengths

Research 
phase

THz 
imaging

I?, II Yes Possible Active Both THz range Research 
phase

IR in spatial  
domain

I, II, III Yes Yes Can be both Remote Naturally  
emitted IR

Research 
phase

CSLM II No Yes Active Contact Different 
wavelengths

Available

OCT II No Yes Active Contact Near-IR Research 
phase

Abbreviations: CSLM, confocal scanning laser microscopy; IR, infrared; MSF, multispectral imaging in the frequency domain; MSS, multispectral imaging in the spatial domain; 
OCT, optical coherence tomography; PMA, premarket approval.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

201

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

Dermoscopy involves the use of a handheld microscope 

and polarized light, which allows imaging of the deeper 

structures of the skin lesions. Reported sensitivity and 

specificity of dermoscopy are 98% and 68%, respectively.12 

Dermoscopy has been reported to improve the diagnostic 

accuracy for melanoma by 5%–30%. However, studies have 

also confirmed that diagnostic accuracy is heavily dependent 

on the operator’s experience,11–14 and both techniques are 

highly subjective, without broadly applicable standards or 

quantitative criteria. A digital photo and a dermoscopy image 

of a melanoma lesion are displayed in Figure 2B and C.

Whole-body photography (also referred to as full-body 

photography), which is essentially white-light, high-resolution 

digital photography, has become an important aid in patient 

self-exams. It involves the taking of photographs of the 

entire body, in order to identify suspicious lesions that may 

be malignant melanomas. This method is frequently used 

for high-risk patients. The patient uses an album with a set 

of professional, high-resolution photographs to compare the 

appearance of the lesion at the time of the self-exam with 

a baseline recorded earlier. Photographs of single lesions 

are used for the monitoring of a particular lesion when 

biopsy does not seem warranted. The expectation is that the 

use of photography combined with visual inspection will 

reduce unnecessary biopsies and allow for early detection 

of melanoma.

As far as insurance coverage is concerned, photographic 

methods (or other imaging techniques) in melanoma 

screening are often still considered to be experimental or 

investigational, since there is insufficient published clinical 

data to support a benefit for high-risk patients, which is 

coupled with the lack of standardization of methods and 

guidelines regarding when a lesion needs to be biopsied as 

opposed to monitored. Many insurance companies claim that 

there is currently a lack of sufficient evidence of improved 

patient outcomes (early detection or reduction in numbers 

of unnecessary biopsies) to justify the additional cost of 

imaging, and so many high-risk patients are forced to pay 

for the imaging (for example, total-body photography) out of 

pocket. However, considering the costs for treating advanced 

melanoma and the outcomes, it seems that encouraging the 

use of imaging in diagnostics would make sense and reduce 

morbidity and mortality as well as health-care costs on the 

long run.

Commercial systems for assessment of lesions using 

digital photography already exist; however, they are not 

in widespread use. Implementing the ABCDE diagnos-

tic criteria in an automated evaluation algorithm on a 

computer is extremely challenging and has not yet been 

accomplished because of the diversity and complexity of 

lesions and images. Any commercial system to be used as 

a diagnostic tool in clinical practice would need extensive 

and expensive validation in clinical trials and FDA approval. 

With enhancements in imaging and computer technologies as 

well as image-processing software, the method of the future 

would be high-resolution 3D mapping of the entire body as 

a periodic screening tool. Ideally, the photographic imaging 

technique would be coupled with another more quantitative 

diagnostic technique.

Multispectral imaging
Spectral methods fall into the class of emerging new 

technologies that are currently being investigated to 

determine their ability to diagnose melanoma. Multispectral 

information can be acquired and analyzed both in the spatial 

domain (detector measures radiation intensity at a particular 

point or region) and the spectral domain (detector captures 

radiation intensity as function of wavelength – spectrum – 

for the selected point or domain). The imaging equipment 

and the signal-analysis methods strongly depend on the 

approach taken. These methods hold the promise of offering 

quantitative criteria for melanoma diagnosis.

Multispectral imaging in the spatial 
domain
One of the most significant advances in the detection of 

melanoma over the past decade has been the development 

of the MelaFind15 technology by Mela Sciences.16 After a 

2-year approval process, this technology received PMA 

(P090012) from the FDA for use in the US in 2011. The PMA 

approval was followed by approval for sale of the system in 

the European Union as well.

MelaFind delivers objective additional information 

(based on quantitative characterization of lesions) about 

clinically atypical cutaneous pigmented skin lesions with 

either clinical or historical characteristics of melanoma 

during skin examinations.17–21 The role of the instrument is 

to aid the decision of the dermatologist regarding diagnosis 

and the need for biopsy. The multicenter blinded study, 

using histologic data as a reference standard, supporting 

the PMA application for MelaFind involved 1383 patients 

in the US (enrolled from January 2007 to July 2008). The 

study demonstrated a sensitivity of over 98% of MelaFind 

(95% lower conf idence bound) with a biopsy ratio of 

10.8:1.22 Most lesions were thin melanomas or borderline 

lesions. The MelaFind outcomes were binary: positive 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

202

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

(lesion should be considered for biopsy) and negative (lesion 

should be considered for later evaluation). MelaFind had 

an average specificity of 9.5%, whereas the specificity of 

the investigators was 3.7%.22 Clinicians participating in the 

study were blinded to the output of MelaFind, and patient 

care was managed based on clinical information. The lesion 

was considered positive when the prebiopsy dermatologic 

diagnosis was either melanoma or melanoma could not be 

ruled out. The study could not determine the true sensitivity 

of the participating dermatologists, since only lesions 

scheduled for biopsy were evaluated; melanomas missed 

by examining clinicians were not included in the study. The 

sensitivity of physicians can be assessed in longitudinal 

(long-term follow-up) studies, and such results were not 

available for this PMA because of the short time frame. Other 

studies list biopsy sensitivity values in the range of 86.7%–

93.7%.22 In addition to this, a pilot study investigated the 

biopsy sensitivity of dermatologists in a reader study using 

25 randomly selected melanomas and 25 non-melanomas. 

The readers were 39 dermatologists who did not participate 

in the clinical trial, and they reviewed the clinical history 

and the images of the lesions to decide whether to biopsy the 

lesion or rule out melanoma. The average biopsy sensitivity 

of the reader study was 78%,22 a result similar to some prior 

studies. More details about the MelaFind study are available 

in the original publication.22

The MelaFind instrument characterizes and classifies 

lesions using three outcomes, based on the degree of 3-D 

morphological disorganization of the lesion. The f irst 

outcome describes “high disorganization” lesions (such 

as malignant melanoma, melanoma in situ, high-grade 

dysplastic nevi, and atypical melanocytic proliferation), 

which are candidates for biopsy. The second outcome is 

“non-evaluable”, and biopsy decision for these is based on 

careful evaluation of other clinical criteria.

The MelaFind scanner is suitable for noninvasive assess-

ment of lesions that are sufficiently pigmented to be clearly 

discerned from the surrounding normal skin using automated 

image-processing tools. Their diameter should be in the range 

between 2 mm and 22 mm, and this dimension is determined 

by the imaging optics. Other restrictions and requirements 

involve accessibility by the handheld component, intact 

skin without scars, fibrosis, or foreign matter. MelaFind is 

not suitable for use on several anatomic sites, such as acral, 

palmar, plantar, mucosal, or subungual areas, as well as 

near the eyes.

The handheld component of the MelaFind imaging 

system consists of a radiation source (illuminator) that 

sequentially irradiates the lesion with 10 wavelengths of 

light (therefore, it is classified as a multispectral system that 

collects information in the spatial domain), including visible 

and near-IR bands. Light scattered back from regions beneath 

the surface of the lesion is collected by a lens system that 

focuses it onto a detector, forming one image for each of the 

wavelengths. The spatial resolution of the detector is around 

3 melanocytes (20 µm). Multispectral data are captured 
from a region up to 2.5 mm deep into the skin, and they 

provide unique 3-D information regarding the morphological 

organization of the lesion. The information collected by 

the detector (ten images) is next analyzed by MelaFind’s 

automatic data-analysis algorithms. The data-processing 

steps include calibration (reduction of noise and artifacts), 

followed by quality control (detection of imaging problems 

such as overexposure, underexposure, inappropriate lesion 

size, too much hair, etc). When problems are detected, the 

system instructs the operator to rescan the lesion.

The data-processing algorithms include the application 

of image-processing tools to the data received by the 

detector, such as lesion segmentation to identify portions 

of the image that belong to the lesion, followed by feature 

extraction to quantify the parameters that characterize lesions. 

Parameters of the lesion, such as asymmetry, color variation, 

and border changes (characteristics similar to the ABCDE 

criteria, evaluated by the computer) are quantified for the ten 

wavelengths used by MelaFind. There are differences in the 

skin’s response (penetration depth, scattering characteristics) 

to the ten wavelengths used, and the differences between 

the images recorded at different wavelengths yield useful 

information for the decision-making process. The collected 

data are compared with information stored in the proprietary 

database of melanomas and benign lesions assembled by 

MelaSciences. The database was developed, trained, and 

tested with over 10,000 in vivo lesions and corresponding 

histological results in over 7000 patients. The sample included 

600 melanomas, most of which were early stage lesions. This 

lesion database evolves and “learns” whenever a new case is 

evaluated and its parameters added to the database. The final 

element of the lesion assessment is classification based on the 

level of disorganization. MelaFind is currently in phase III 

clinical trials, and the technology holds the promise of being 

incorporated into clinical practice after further research and 

evaluation.

MelaFind is a type II instrument according to the 

classification introduced in this review, and early clinical 

studies are very promising.22 MelaFind also holds the promise 

of potentially becoming a screening device (type I in this 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

203

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

review) after being proven as a type II device and appropriate 

modifications. Since the method requires irradiation by ten 

wavelengths, implementation as a type III scanner poses 

significant challenges, and it is less likely to prove feasible for 

full-body screening applications. Being recognized as a test 

method by insurance carriers would be a key step towards using 

diagnostic instruments to detect early melanoma, and extensive 

clinical validation will be needed to achieve this goal.

Multispectral imaging in the frequency 
domain: spectroscopic methods
Multispectral methods (in the spectral domain with radiation 

data for a point or region presented as intensity as a function 

of frequency or wavelength) are widely used in engineering 

for the characterization of materials, and they have also 

found their way into ex vivo and more recently into in vivo 

medical diagnostics. In this review advances and important 

features of IR and Raman spectroscopy as representatives 

of spectroscopic methods are summarized.

IR and Raman spectroscopy provide details regarding 

the chemical composition and molecular structure of 

substances in cells and biological tissues, and they are 

considered to be vibrational spectroscopic techniques.23,24 

Over time, with the progress of technology, they have also 

evolved into effective visualization tools. Since pathological 

changes and diseases cause chemical and structural changes 

in tissues on the molecular level, the resulting differences 

in the vibrational spectra serve as sensitive markers or 

unique fingerprints of the disease.25 These spectra can be 

captured by spectroscopic methods, which are noninvasive 

and do not require the administration of contrast. Early 

biomedical applications of these two spectroscopic methods 

would allow the measurement of spectra at a particular 

point in the tissue, and required therefore knowledge of 

the exact location of diseased tissue. With advances in 

optical imaging systems over the past 15 years, sensitive, 

high-throughput instruments have become available. These 

enhanced instruments collect a large number of spectra from 

larger samples faster and with improved spatial resolution 

with the aid of a microscopic imaging system. Also, over 

the past decade, fiber-optic probes have been developed, 

and their feasibility for in vivo applications has been 

demonstrated. Both spectra – the IR absorption spectrum 

and the Raman scattering spectrum – carry the same type of 

information regarding the energies of molecular vibrations. 

The underlying physics for these two spectroscopic methods 

is different, and each method offers specific advantages (and 

disadvantages) in medical applications.

IR spectroscopy
IR spectroscopy measures absorbed radiation, and can serve 

as a visualization tool to aid the pathologist in evaluating 

tissue specimens. IR instrumentation is usually less complex 

than devices for Raman spectroscopy that measure scattered 

light. Collection of IR data is faster, and signal-to-noise ratios 

are higher than for Raman spectrometers. The disadvantage 

of the method is that sample thickness for aqueous biological 

systems is limited to a few micrometers in transmission 

or attenuated total reflection (ATR) measurements. This 

limitation is imposed by water being a strong IR-absorbing 

medium. These samples are prepared on special substrates, 

and it is important to avoid affecting the composition of the 

sample during the preparation. Advanced Fourier transform 

IR (FTIR) spectrometers allow two-dimensional (2D) and 3D 

mapping of relevant spectral characteristics of the tissue. In 

IR spectroscopic imaging, the field of view is irradiated and 

imaged onto a detector array whose elements are sensitive to 

IR radiation. Each sensor element in the array will collect a 

spectrum, and the number of spectra collected simultaneously 

depends on the number of sensor elements in the array. IR 

radiation can be guided to the sample, and the reflected or 

transmitted radiation routed to the spectrometer, by means 

of IR fiber optics. Fiber-optic systems are suitable for in vivo 

endoscopic measurements; however, contact between the 

tissue and probe and probe contamination remain challenges 

to be resolved in the future.

Raman spectroscopy
Since the presence of water affects Raman spectra little, 

Raman spectroscopy is better suited for medical diagnostic 

applications. Raman spectra are typically measured in 

the reflection mode, and Raman spectroscopy can also be 

applied in vivo. Penetration of light can vary, depending on 

the wavelength, power, and sample properties (geometry 

and composition).

Human skin is a multilayered system, with layer 

thicknesses in the 10–500 µm range. Because of the 
absorption of IR radiation by water, the depth of penetration 

of IR radiation in skin tissue in in vivo IR spectroscopy 

measurements is only a few micrometers. Therefore, the ATR 

mode of spectroscopy can be used to analyze the outermost 

layer of stratum corneum.26 FTIR spectroscopy has also 

been used to analyze healthy skin and lesions.27–29 The ATR 

technique coupled to IR fiber optics has been applied for 

FTIR imaging of the stratum corneum to evaluate intact, 

thin skin sections, and the effects of exogenous materials.30,31 

FTIR spectroscopy has been combined with Raman imaging 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

204

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

to analyze the stratum corneum and the permeation of lipids 

into the skin.32–34 Confocal Raman microspectroscopy has also 

been used to study structures in the stratum corneum35–37 in up 

to 200-µm depth. The combination of Raman spectroscopy 
and Confocal Scanning Laser Microscopy (CSLM) offers 

the ability to analyze sections and layers of the skin without 

physically dissecting the tissue.38 The feasibility of these 

methods for the diagnosis of melanoma and other skin 

cancers has been demonstrated in several studies.39–43

Spectroscopic methods are very promising tools for 

the early diagnosis of melanoma. In addition to ex vivo 

characterization of diseased tissue, current research is 

focusing on in vivo applications. After adequate calibration 

with a large number of samples, automated evaluation of 

the data should be possible. Spectroscopic systems are best 

suited for type II devices, and a simplified instrument could 

possibly be converted into a type I device for screening 

purposes. Since the investigated tissue has to be irradiated 

with EM radiation of appropriate wavelength, it would be 

hard to accomplish both irradiation and the capturing of the 

scattered and reflected radiation for a large surface area in 

full-body measurements in type III devices.

Terahertz imaging
THz (1 THz – 1012 Hz) imaging uses radiation whose 

frequency lies between the microwave and IR regions 

(longer than IR wavelengths) of the EM spectrum. Until 

recently, it was the least explored and utilized region of 

the spectrum, due to difficulties involved in building THz 

radiation sources and detectors. Advances in optics and 

instrumentation over the past two decades have led to the 

development of a range of sources and detectors. One of 

the many applications of THz imaging is in the detection 

of disease, in particular skin cancer. The aim is to develop 

small, portable, fast, handheld scanners for early detection 

of cancer. The technology has shown promise; however, 

more research and development is needed to develop an 

imaging system capable of reliable automated detection 

of melanoma.

Infrared imaging in the spatial domain: 
infrared thermography revisited
The advances in IR (thermographic) imaging technology 

that captures spatial IR-emission data (rather than spectral 

characteristics), computers, and image-analysis techniques 

over the past two decades has led to renewed interest in 

IR imaging, in particular quantitative imaging, in a variety 

of medical and engineering applications. Since cancerous 

lesions are metabolically more active than healthy tissue and 

blood supply to the lesion is increased, cancerous lesions 

are warmer, and under certain conditions the difference in 

the IR radiation emitted by the lesion and healthy tissue can 

be readily detected by thermographic imaging.

Thermal imaging has been used as a research tool in 

medicine for over 50 years44–47 to analyze medical conditions 

associated with changes in the body or skin temperature. 

Skin cancer and melanoma detection are some of the  earliest 

 applications of IR imaging, and the value of the method has 

been a subject of controversy in the past. However, recent 

studies with advanced instrumentation and image-analysis 

tools have shown characteristic differences in thermal 

signatures between healthy skin or benign lesions and 

 cancerous lesions.48–50 Some new results and methods will be 

discussed in this review. In this section, we highlight general 

issues and challenges in thermographic imaging as well as our 

recent efforts in quantitative thermal imaging of melanoma.

Thermographic techniques capture the EM radiation 

naturally emitted by the subject, the human body, and this 

radiation closely resembles the black-body radiation at 

36°C–37°C; its spectrum is shown in Figure 1 (top right). 
The emissivity of the skin is in the range 0.96–0.98. The 

capturing of naturally emitted radiation is an advantage over 

methods that require irradiation of the skin by an external 

source (to measure the reflected or scattered radiation in 

order to analyze skin structures) such as by MelaFind, in 

particular for full-body imaging applications. Therefore, 

the method is inherently safe and noninvasive. The emitted 

radiation is collected by the imaging optics and focused onto 

the focal plane array sensitive to IR wavelengths, thereby 

delivering the information in spatial (contrasted to spectral 

information in spectroscopic methods) and two dimensional 

(2-D) format. Recently, high-sensitivity uncooled detectors 

called focal plane arrays have become available, which 

have greatly simplified the technology, with the promise of 

significant cost reductions in the future. Another favorable 

feature of the method is that it is both image-forming (it 

allows the visual mapping of thermal structures on the skin 

surface and underneath the skin) and quantitative (allows 

measurement of temperature, metabolic activity rates, and 

other thermophysical properties based on temperature data). 

Depending on the number of detectors in the focal plane–

array matrix, a single image can contain 2D information 

with thousands of temperature measurements recorded in a 

fraction of a second. Modern IR cameras offer excellent time 

resolution, so the temperature changes of the subject can be 

recorded over time in the form of a movie. Images can be 

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

205

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

displayed as gray-scale or color coded with a color palette. 

The rainbow palette is preferred in medical applications, with 

the blue corresponding to low and the red to high temperature. 

Assigning absolute temperatures to emitted radiation poses 

a significant challenge and requires careful calibration of 

the instrument, as well as standardization of the imaging 

process. In addition to the visual information in the form 

of a 2D image or movie, quantitative information about the 

temperature evolution of a point or a region over time can be 

extracted from the 2-D information by postprocessing.

The value of thermographic imaging goes beyond 

providing visual insight into the thermal characteristics 

of the skin and the human body: proper processing of 

the temperature data offers the possibility of quantitative 

comparison and automated detection and quantification 

of abnormal events or processes or establishing other 

associations with medical conditions of interest.51–55 A good 

understanding of the thermal physiology of the body is 

essential for the interpretation of temperature data and 

spatiotemporal temperature changes in thermographic 

images. To enhance the understanding and interpretation 

of the information contained in thermographic images, 

computational modeling of the physiological processes has 

to go hand in hand with the thermographic imaging steps.52 

When applying thermography in medical diagnostics, the 

voluntary and involuntary movement of the subject during 

the imaging process introduces an additional complexity 

into the image-analysis and quantification process. While 

specific image features are easy to detect and track in a 

series of images recorded over time in white-light imaging, 

the same features are not readily visible in the IR image 

sequence. Therefore, motion-tracking procedures developed 

specif ically for IR imaging are needed for quantitative 

thermographic imaging.

IR imaging can be implemented either as a passive (static) 

or an active (dynamic) visualization method. Active IR 

imaging involves introducing external forcing, such as heat-

ing, cooling (also called thermostimulation in breast cancer 

imaging), or pressure to induce or enhance relevant thermal 

contrasts between the investigated lesion and healthy tissue. 

The general advantage of active measurement methods is that 

they do not require the subject to reach a well-defined steady 

state, which can be very time-consuming and challenging. 

For example, in the past, when static applications prevailed, 

the patient was required to spend one or more hours in a 

thermally conditioned exam room in order to fully accli-

mate (reach a thermal steady state) to the environment. 

Reaching a true steady state may not always be possible, 

and individual variations between test subjects will remain. 

Therefore, standardization in the imaging process – the key 

to comparison of images recorded at different times or of 

different subjects – was and remains a significant challenge 

in passive imaging.

In the dynamic method, the thermal response to an 

excitation, often applied as a step change in the thermal 

boundary conditions, involving the application or removal 

of heating or cooling, is measured. This is a much faster and 

more robust approach than the passive method. The dynamic 

method is well established in numerous engineering tests and 

applications, such as thermophysical property measurements, 

and it offers advantages in clinical applications, for which the 

duration of the measurement and the ease of use are critical. It 

allows the detection of abnormal features by self-referencing, 

ie, comparing the responses of healthy and diseased tissue of 

the same subject, rather than comparing responses of different 

subjects. While the application of external forcing increases 

the complexity of the measurement, the benefits outweigh 

the drawbacks. The required heating or cooling has to induce 

small changes in skin temperature; therefore, the duration 

of cooling is short and the temperature levels applied do not 

cause significant discomfort to the patient.

When the skin surface is cooled in active IR imaging, 

the difference in the thermophysical properties of the lesion 

(located near or underneath the surface) when compared to 

healthy tissue results in identifiable temperature contours 

and temperature differences between the lesion and the 

surrounding healthy tissue during the thermal recovery phase. 

These surface-temperature contours will differ from those 

present in the steady-state situation observed during passive 

IR imaging. Therefore, the measurement process consists of 

three phases: (1) the initial phase, when the skin is exposed 

to nominal ambient conditions, (2) the cooling phase, which 

is followed by (3) the thermal recovery phase.

In the Heat Transfer Lab of Johns Hopkins University, 

we developed a thermographic imaging system that 

allows accurate measurements of transient skin-surface 

temperature distributions51,53–57 and can be used for the 

detection of melanoma. It relies on active (dynamic) 

IR imaging based on cooling for the characterization of 

the thermal response of skin lesions and healthy skin. 

Cancerous lesions, such as melanomas, generate more heat 

and reheat faster than the surrounding healthy skin, thereby 

creating a marker for melanoma detection. The aim of the 

imaging is to distinguish benign lesions (which behave 

thermally similar to the healthy skin) from malignant 

pigmented lesions.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

206

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

The steps in the imaging process are illustrated in Figure 3. 

The imaging starts with applying a rectangular template to 

the region of interest, centered around the lesion. A white-

light photograph of the domain is taken first, followed by 

an IR image of the reference steady-state situation under 

ambient conditions (Figure 3A). The lesion is not visible in 

the reference steady-state IR image, implying that the heat 

generation is too small to be measured at the skin surface 

under static conditions. The skin is then cooled by blowing 

cold air or applying a cooling patch at 15°C–25°C for a 
duration of up to 1 minute. After removing the cooling load, 

the dynamic thermal response of the structure is acquired 

using IR imaging. Color-coded IR images recorded 2 seconds 

into the thermal recovery are shown in Figure 3B. The results 

suggest that thermal contrast between the lesion and the 

healthy surrounding tissue is enhanced by the cooling. After 

the postprocessing steps, the temperature of the lesion and 

healthy tissue are plotted as a function of time during the 

thermal recovery, as shown in Figure 3C. There is a significant 

difference in temperature between the melanoma lesion 

and healthy tissue during thermal recovery, whereas benign 

pigmented lesions have the same thermal recovery as the 

healthy tissue. The temperature difference can be measured 

accurately using modern IR cameras. More details about 

the method are available in the literature.51–57 This imaging 

technique holds the promise of staging melanomas based on 

the magnitude of temperature differences and other thermal 

characteristics of the lesion during the recovery process.

Motion tracking in dynamic  
IR imaging
IR imaging is a noncontact method, and this feature offers 

advantages in terms of ease of application and the ability to 

image larger surface areas and multiple lesions. Therefore, 

this method is suitable for type III full-body imaging 

systems. However, during the thermal recovery phase, 

involuntary movement of the patient is unavoidable in a 

clinical environment, and even breathing can cause small 

spatial displacement of lesions that can lead to deterioration 

of the measurement data. The measurement involves accurate 

tracing of the transient temperature response at any specific 

point on the skin. To accomplish this, it is necessary to apply 

motion tracking/compensation processing on the IR video 

sequence. The framework of motion tracking can be generally 

summarized in two steps.57

Step 1: registration between the white-
light image and the first IR image frame
Since the lesion is visible in the white-light image only, to 

identify its corresponding points in the IR image, the first 

task is the registration via feature point correspondence. 

An adhesive marker, visible both in white-light and IR 

images, is used for the registration computations. As 

shown in Figure 4A, the four corners of the marker are first 

identified in the white-light image either manually or by 

an edge-detection algorithm, such as the Harris method.58 

Next, the corresponding points are identified in the first IR 

image frame, which serves as the reference image in step 2 

(Figure 4B). Based on the four pairs of points for coordinate 

correspondence, a 2-D projective transformation matrix can 

be identified to map any point in the white-light image to 

its corresponding point in the reference IR image frame.59 

The lesion region can be delineated in the white-light 

image via an interactive segmentation algorithm, such as a 

250

200

150

100

Recovery: 2s

Lesion

Steady-state

Lesion

50

10

20

30

40

50

60

70

80

60 70 80 90 100 110 120 130 140 150

18

35

34

33

32

31

30

29

28

27

26

25

17

16

15

14

13

12

11

10

50 100 150 200 250 300

30

30025020015010050
250

200

150

100

50

25

20

15

10

C

0 60 120 180
10

15

20

25

30

35

t (s)

T
 (

°C
)

lesion

healthy skin

C

B

A

0 60 120 180
10

15

20

25

30

35

t (s)

Lesion

Healthy skin

Figure 3 (A) white light photograph of the larger body surface area with a cluster 
of pigmented lesions, adhesive window serving as thermal marker and reference IR 
image of the region at ambient temperature; (B) the same area 2s into the thermal 
recovery and magnified section of the melanoma lesion and its surroundings; (C) 
temperature profiles of the lesion and the surrounding normal skin during the 
thermal recovery process.51–53

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

207

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

random walker.60 Once the transformation matrix is solved, 

the delineated lesion can be transferred to the registered IR 

image (Figure 4C).

Step 2: registration between  
the consecutive IR image frames  
in the recovery IR video sequence
After the lesion location is identified in the first IR image 

frame, registration can be applied between every consecutive 

IR frame in the video sequence. Any point of interest can then 

be tracked in the video in the presence of involuntary subject 

movement. The registration between the consecutive IR 

frames can be achieved using a quadratic motion model.61

Transformations of the image during the motion-

correction steps are illustrated in Figure 5. The magnitude 

and direction of the motion are indicated as a vector in each 

image frame. A sequence of IR image frames recorded 

during the recovery phase can be aligned to compensate for 

involuntary body/limb movement of the patient by applying 

a quadratic motion model for registration. The registration 

of the lesion region enables an accurate comparison between 

the transient thermal response of healthy skin and that of 

the lesion, and provides critical information that allows 

identification of the malignant lesion, as shown by Pirtini 

Cetingul and Herman.51,52,54 Since IR imaging (with proper 

calibration) yields quantitative data, the method (analysis of 

the thermal recovery process) holds the potential of allowing 

the staging of the disease.

This version of quantitative thermography has the 

potential to be implemented as a type I, type II, and type III 

imaging device, since the radiation naturally emitted by 

the body is measured, and direct contact between the skin 

and the probe is not required. The drawback is the need for 

sophisticated motion tracking in the image-processing phase 

to compensate for involuntary movement of the subject. IR 

thermography alone cannot distinguish between different 

types of skin cancer; however, it can alert the patient to 

the need to seek out professional help (type I device for 

screening). As a type II device, with proper calibration, it 

could potentially measure the magnitude of excess heating 

(caused by increased metabolic activity and blood supply 

in the cancerous lesion) and provide a quantitative measure 

of the malignant potential. The discipline of IR thermogra-

phy has been experiencing dramatic growth over the past 

decade, primarily driven by night vision and surveillance 

applications, accompanied by improvements in hardware 

and reduction in cost of the instrumentation. With current 

trends continuing, thermography would be a good option for 

large-scale screening efforts at the level of the patient or in 

primary care facilities.

Confocal scanning laser microscopy
CSLM is a noninvasive imaging method that has the ability to 

assess the cellular and nuclear details of skin lesions in vivo, 

with details similar to histology.62,63 CSLM detects backscat-

tered light, and contrast is caused by natural variations of the 

refractive index of tissue microstructures,64 such as organelles 

and melanosomes. Since the skin pigment melanin has a high 

Without
correction

With
correction

Without
correction

With
correction

0 s

14 s

36 s

80 s

124 s

148 s

Figure 5 Motion correction for compensation of the subject’s involuntary 
movement in IR images recorded at different time instants. 
Notes: The effects of motion correction can be observed in the corrected columns 
as dark regions around the edges of the images. white arrows in the uncorrected 
images represent the direction and magnitude of motion of the subject.

30

25

20

15

(1997, 2508)
(1047, 2458)

(1182, 1406)
*

(2130, 1550)
*

(1637, 1968)

(1482, 1706)
*

**

A B C

Figure 4 Image registration steps in melanoma detection: (A) white light image with characteristic points for image analysis (rectangular marker and lesion center), (B) edge 
and corner correspondence in IR image, (C) lesion contour registration in the IR images via transformation matrix.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

208

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

refractive index, it behaves as a contrast agent for cytoplasm.65 

Therefore, the cytoplasm of melanocytes in pigmented and 

amelanotic melanomas appears brighter and can be easily 

detected. Past small-scale clinical studies focused on the 

detection of skin cancers, such as melanomas,66–69 basal and 

squamous cell carcinomas,70 benign lesions, and the detec-

tion of margins between lesions and surrounding normal 

tissue.71 Clinical studies have shown that the technique has a 

sensitivity in the range of 89%–97% and a specificity in the 

range of 80%–85% (depending on the study),62–71 after appro-

priate training of the physician in the interpretation of the 

patterns observed using CSLM. Similar to other diagnostic 

methods that involve interpretation of visual images, the 

diagnostic accuracy will be dependent on the experience of 

the operator.

A commercially available CSLM imaging system 

(VivaScope) is shown in Figure 7A. The method is capable of 

assessing parallel layers of the skin up to a depth of 300 µm 
(Figure 7B). The head of the imaging system contains the 

optical components (light source and imaging optics) as 

well as a computer-controlled linear translation stage for 

sequential scanning of the area of interest. To minimize 

0 50 100 150 200
10

15

20

25

30

35

Time (sec)
0 50 100 150 200

Time (sec)

T
e
m

p
e
ra

tu
re

 (
°C

) 
10

15

20

25

30

35

T
e
m

p
e
ra

tu
re

 (
°C

) 

Recovery temperature (without motion correction)

Lesion

Healthy tissue

Recovery temperature (with motion correction)

Healthy tissue

Lesion

Figure 6 Impact of the motion correction on the reconstructed temperature profiles. 
Notes: without motion correction (left), the noise is of the order of magnitude of the measured signal. with motion correction. the temperature of the melanoma lesion is 
higher than the temperature of the surrounding healthy tissue. This temperature difference can be used to quantify the malignant potential of the cancerous lesion.

25 µm

25 µm

25 µm

25 µm

Without a mole

Stratum corneum
Stratum granulosum

Stratum spinosum Dermo-epidermaljunc.

A

B

C

Confocal microscope
field of view

2 mm

2 m
m 20 µm

x

y

z

Figure 7 (A) vivaScope imaging system for dermatology; (B) parallel skin layers imaged by the CSLM system; (C) CSLM images of different skin layers.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

209

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

optical losses between media with very different refractive 

indices, an immersion fluid is applied between the skin and 

the optics to enable direct contact between detector and skin. 

The positioning of the detector head relative to the lesion may 

pose a challenge for some lesion locations. A CSLM system 

and images of four skin layers are displayed in Figure 7, 

and they deliver fine detail and high-quality information on 

skin structures.

CSLM is a commercially available type II imaging system 

suitable for use in specialized centers by trained operators. 

Since it requires direct contact between optics and the skin 

by means of an immersion fluid, this imaging system is not 

suitable for a full-body scanner. Also, since the interpretation 

of the images requires specialized knowledge and automated 

interpretation of the images would pose a major challenge, it 

is not likely that CSLM will be suitable as a type I screening 

instrument.

Optical coherence tomography
OCT is a noninvasive, in vivo imaging method, which captures 

high-resolution (µm), 3D images of biological tissue. OCT is 
an interferometric technique using relatively long-wavelength 

light in the near-IR portion of the spectrum, which is able to 

penetrate into the scattering medium (deeper than CSLM). This 

technique has higher resolution, greater detection depth and 

image size, and faster scanning time than CSLM. Although in 

OCT, a melanoma shows increased light scattering and more 

homogenous signal distribution than healthy skin, more work is 

needed to improve its utility for skin cancer detection.12,13 OCT 

is suitable as a type II device for use in specialized centers by 

operators trained in the interpretation of the images.

Conclusions
This brief review discusses some of the promising current 

technologies as well as the needs and challenges in 

developing sensitive and reliable diagnostic tools for the 

early detection of melanoma. Melanoma is the fastest-

growing cancer in terms of incidence, and the need for 

accurate diagnostic tools is increasing. Every year, around 

2.5–3 million skin lesions are evaluated in the US, and 

over 100,000 are diagnosed as melanoma. The objective is 

to develop automated diagnostic instruments for screening 

of individual lesions and full-body screening, as well as 

sophisticated instruments that can provide dermatologists 

with fine detail regarding the structure of a lesion and 

staging information in vivo. Screening instruments would 

alert patients to seek the care of a dermatologist, and 

would be intended for use in primary care facilities or by 

the patients, similar to blood pressure cuffs or diabetes-

testing equipment. With recent progress in electronics and 

instrumentation, several sophisticated and very promising 

imaging methods have emerged and are being investigated in 

small trials. One of the key challenges is that diagnostic 

instruments are expected to compete in price and ease of 

use with visual inspection, which is the current standard of 

care. Clearly, using a diagnostic instrument would increase 

the duration and the cost of the exam; therefore, insurance 

coverage would be a key driving factor for technology 

development. The funding for technology development 

beyond the initial feasibility studies and the funding of large 

scale studies to demonstrate the effectiveness of imaging 

systems along with the complex and lengthy governmental 

approval process are the main challenges on the path of these 

imaging systems finding their place in medical care.

Acknowledgments
Images and photographs for this review were taken or 

contributed by Rajeev Hatwar, Tze-Yuan Cheng, and Muge 

Pirtini Cetingul. This research was funded by the National 

Institutes of Health NCI (grant no 5R01CA161265-02), the 

National Science Foundation (grant no 0651981), and the 

Alexander and Margaret Stewart Trust through the Cancer 

Center of Johns Hopkins University.

Disclosure
The author reports no conflicts of interest in this work.

References
1. National Cancer Institute. SEER Stat Fact Sheets: Melanoma of the Skin. 

2012. Available from: http://seer.cancer.gov/statfacts/html/melan.html. 
Accessed August 3, 2012.

2. Skin Cancer Foundation. Skin Cancer Facts. 2012. Available from: http://
www.skincancer.org/Skin-Cancer-Facts. Accessed August 3, 2012.

3. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, 
2009. CA Cancer J Clin. 2009;59(4):225–249.

4. Elder D. Tumor progression, early diagnosis and prognosis of melanoma. 
Acta Oncol. 1999;38(5):535–547.

5. Fecher LA, Cummings SD, Keefe MJ, Alani RM. Toward a molecular 
classification of melanoma. J Clin Oncol. 2007;25(12):1606–1620.

6. Wartman D, Weinstock M. Are we overemphasizing sun avoidance in 
protection from melanoma? Cancer Epidemiol Biomarkers Prev. 2008; 
17(3):469–470.

7. Geller AC, Swetter SM, Brooks K, Demierre MF, Yaroch AL. Screening, 
early detection, and trends for melanoma: current status (2000–2006) 
and future directions. J Am Acad Dermatol. 2007;57(4):555–572; quiz 
573–576.

8. Friedman RJ, Rigel DS, Kopf AW. Early detection of malignant 
melanoma: the role of physician examination and self-examination of 
the skin. CA Cancer J Clin. 1985;35(3):130–151.

9. Abbasi NR, Shaw HM, Rigel DS, et al. Early diagnosis of cutaneous 
melanoma: revisiting the ABCD criteria. JAMA. 2004;292(22): 
2771–2776.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

210

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

http://seer.cancer.gov/statfacts/html/melan.html
http://www.skincancer.org/Skin-Cancer-Facts
http://www.skincancer.org/Skin-Cancer-Facts
www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology 2012:5

 10. Thomas L, Tranchand P, Berard F, Secchi T, Colin C, Moulin G. Semio-
logical value of ABCDE criteria in the diagnosis of cutaneous pigmented 
tumors. Dermatology. 1998;197(1):11–17.

 11. Wang SQ, Rabinovitz H, Kopf AW, Oliviero M. Current technologies 
for the in vivo diagnosis of cutaneous melanomas. Clin Dermatol.  
2004;22(3):217–222.

 12. Psaty EL, Halpern AC. Current and emerging technologies in melanoma 
diagnosis: the state of the art. Clin Dermatol. 2009;27(1):35–45.

 13. Patel JK, Konda S, Perez OA, Amini S, Elgart G, Berman B. Newer 
technologies/techniques and tools in the diagnosis of melanoma. Eur 
J Dermatol. 2008;18(6):617–631.

 14. Andreassi M, Andreassi L. Utility and limits of noninvasive methods 
in dermatology. Expert Rev Dermatol. 2007;2(3):249–255.

 15. http://www.melafind.com.
 16. http://www.melasciences.com.
 17. Gutkowicz-Krusin D, Elbaum M, Jacobs A, et al. Precision of 

automatic measurements of pigmented skin lesion parameters with 
a MelaFind(TM) multispectral digital dermoscope. Melanoma Res. 
2000;10(6):563–570.

 18. Elbaum M, Kopf AW, Rabinovitz HS, et al. Automatic differentiation of 
melanoma from melanocytic nevi with multispectral digital dermoscopy: 
a feasibility study. J Am Acad Dermatol. 2001;44(2):207–218.

 19. Elbaum M. Computer-aided melanoma diagnosis. Dermatol Clin. 
2002;20(4):735–747, x–xi.

 20. Elbaum M. Automated diagnosis: illustrated by the MelaFind system.  
In: Marghoob AA, Kopf AW, Braun R, editors. An Atlas of Dermoscopy. 
Abingdon: Taylor and Francis; 2004:325–341.

 21. Friedman RJ, Gutkowicz-Krusin D, Farber MJ, et al. The diagnostic 
performance of expert dermoscopists vs a computer-vision system on 
small-diameter melanomas. Arch Dermatol. 2008;144(4):476–482.

 22. Monheit G, Cognetta AB, Ferris L, et al. The performance of MelaFind:  
a prospective multicenter study. Arch Dermatol. 2011;147(2):188–194.

 23. Rigel DS, Roy M, Yoo J, Cockerell CJ, Robinson JK, White R. Impact 
of guidance from a computer-aided multispectral digital skin lesion 
analysis device on decision to biopsy lesions clinically suggestive of 
melanoma. Arch Dermatol. 2012;148(4):541–543.

 24. Krafft C, Sergo V. Biomedical applications of Raman and infrared 
spectroscopy to diagnose tissues. Spectroscopy. 2006;20(5–6): 
195–218.

 25. Garidel P. Insights in the biochemical composition of skin as 
investigated by micro infrared spectroscopic imaging. Phys Chem  
Chem Phys. 2003;5:2673–2679.

 26. Lucassen GW, Caspers PJ, Puppels GJ. In vivo infrared and Raman 
spectroscopy of stratum corneum. Proc SPIE. 1998;3257:52–60.

 27. Mendelsohn R, Flach CR, Moore DJ. Determination of molecular 
conformation and permeation in skin via IR spectroscopy, microscopy 
and imaging. Biochim Biophys Acta. 2006;1758(7):923–933.

 28. Mendelsohn R, Rerek ME, Moore DJ. Infrared spectroscopy and 
microscopic imaging of stratum corneum models and skin. Phys Chem 
Chem Phys. 2000;2:4651–4657.

 29. Mendelsohn R, Chen HC, Rerek ME, Moore DJ. Infrared microscopic 
imaging maps the spatial distribution of exogenous molecules in skin. 
J Biomed Opt. 2003;8(2):185–190.

 30. Xiao C, Moore DJ, Flach CR, Mendelsohn R. Permeation of dimyris-
toylphosphatidylcholine into skin – structural and spatial information 
from IR and Raman microscopic imaging. Vib Spectrosc. 2005;38(1–2): 
151–158.

 31. Xiao C, Moore DJ, Rerek ME, Flach CR, Mendelsohn R. Feasibility 
of tracking phospholipid permeation from IR and Raman microscopic 
imaging. J Invest Dermatol. 2005;124(3):622–632.

 32. Zhang G, Moore DJ, Mendelsohn R, Flach CR. Vibrational 
microspectroscopy and imaging of molecular composition and 
structure during human corneocyte maturation. J Invest Dermatol.  
2006;126(5):1088–1094.

 33. Bommannan D, Potts RO, Guy RH. Examination of stratum corneum 
barrier function in vivo by infrared spectroscopy. J Invest Dermatol. 
1990;95(4):403–408.

 34. Bhargava R, Levin IW. Gram-Schmidt orthogonalization for rapid 
reconstruction of Fourier transform infrared spectroscopic imaging 
data. Appl Spectrosc. 2004;58(8):995–1000.

 35. Xiao C, Flach CR, Marcott C, Mendelsohn R. Uncertainties in depth 
determination and comparison of multivariate with univariate analysis 
in confocal Raman studies of a laminated polymer and skin. Appl 
Spectrosc. 2004;58(4):382–389.

 36. Caspers PJ, Lucassen GW, Carter EA, Bruining HA, Puppels GJ.  
In vivo confocal Raman microspectroscopy of the skin: noninvasive 
determination of molecular concentration profiles. J Invest Dermatol. 
2001;116(3):434–442.

 37. Caspers PJ, Lucassen GW, Puppels GJ. Combined in vivo  confocal 
Raman spectroscopy and confocal microscopy of human skin. 
Biophys J. 2003;85(1):572–580.

 38. Percot A, Lafleur M. Direct observation of domains in model stratum 
corneum lipid mixtures by Raman microspectroscopy. Biophys J.  
2001;81(4):2144–2153.

 39. Tfalyli A, Piot O, Durlach A, Bernard P, Manfait M. Discriminating 
nevus and melanoma on paraffin-embedded skin biopsies using FTIR 
microspectroscopy. Biochim Biophys Acta. 2005;1724(3):262–269.

 40. Lasch P, Naumann D. FT-IR microspectroscopic imaging of human 
carcinoma thin sections based on pattern recognition techniques. Cell 
Mol Biol (Noisy-le-grand). 1998;44(1):189–202.

 41. Gniadecka M, Philipsen PA, Sigurdsson S, et al. Melanoma diagnosis 
by Raman spectroscopy and neural networks: structure alteration 
in proteins and lipids in intact cancer tissue. J Invest Dermatol. 
2004;122(2):443–449.

 42. Nijssen A, Bakker Schut TC, Heule F, et al. Discriminating basal cell 
carcinoma from its surrounding tissue by Raman spectroscopy. J Invest 
Dermatol. 2002;119(1):64–69.

 43. Huang Z, Lui H, Chen XK, Alajlan A, McLean DI, Zeng H. Raman 
spectroscopy of in vivo cutaneous melanin. J Biomed Opt. 2004;9(6): 
1198–1205.

 44. Cristofolini M, Piscioli F, Valdagni C, Della Selva A. Correlations 
between thermography and morphology of primary cutaneous malignant 
melanomas. Acta Thermogr. 1976;1(1):3–11.

 45. Hartmann M, Kunze J, Friedel S. Telethermography in the diagnostic 
and management of malignant melanomas. J Dermatol Surg Oncol. 
1981;7(3):213–218.

 46. Di Carlo A. Thermography and the possibilities for its applications in 
clinical and experimental dermatology. Clin Dermatol. 1995;13(4): 
329–336.

 47. Diakides NA. 1998 Infrared imaging: an emerging technology in 
medicine. IEEE Eng Med Biol Mag. 1998;17(4):17–18.

 48. Fauci M, Breiter R, Cabanski W, et al. Medical infrared imaging – 
differentiating facts from fiction, and the impact of high precision 
quantum well infrared photodetector camera systems, and other factors, 
in its reemergence. Infrared Phys. 2001;42(3–5):337–344.

 49. Gulyaev YV, Markov AG, Koreneva LG, Zakharov PV. Dynamical 
infrared thermography in humans. IEEE Eng Med Biol Mag. 1995;14: 
766–771.

 50. Ring EFJ, Ammer K. Infrared thermal imaging in medicine. Physiol 
Meas. 2012;33(3):R33–R46.

 51. Herman C, Pirtini Cetingul M. Quantitative visualization and detection  
of skin cancer using dynamic thermal imaging. J Vis Exp. 2011;(51): 
e2679.

 52. Pirtini Cetingul M, Herman C. Heat transfer model of skin tissue for the  
detection of lesions: sensitivity analysis. Phys Med Biol. 2010;55(19): 
5933–5951.

 53. Pirtini Cetingul M, Herman C. Quantification of the thermal signature 
of a melanoma lesion. Int J Therm Sci. 2011;50(4):421–431.

 54. Pirtini Cetingul M, Herman C. The assessment of melanoma risk 
using the dynamic infrared imaging technique. J Therm Sci Eng Appl. 
2011;3(3):031006.1–031006.9.

 55. Pirtini Cetingul M, Alani RM, Herman C. Quantitative evaluation of skin 
lesions using transient thermal imaging. Proceedings of the International Heat 
Transfer Conference, IHTC14; August 8–13, 2010; Washington, DC, USA.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

211

Technologies for detection of melanoma
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

www.melafind.com
www.melasciences.com
www.dovepress.com
www.dovepress.com
www.dovepress.com


Clinical, Cosmetic and Investigational Dermatology

Publish your work in this journal

Submit your manuscript here: http://www.dovepress.com/clinical-cosmetic-and-investigational-dermatology-journal

Clinical, Cosmetic and Investigational Dermatology is an interna-
tional, peer-reviewed, open access, online journal that focuses on 
the latest clinical and experimental research in all aspects of skin 
disease and cosmetic interventions. All areas of dermatology will 
be covered; contributions will be welcomed from all clinicians and 

basic science researchers globally. This journal is indexed on CAS. 
The manuscript management system is completely online and includes 
a very quick and fair peer-review system, which is all easy to use. 
Visit http://www.dovepress.com/testimonials.php to read real quotes 
from published authors.

Clinical, Cosmetic and Investigational Dermatology 2012:5

 56. Pirtini Cetingul M, Alani RM, Herman C. Detection of skin cancer 
using skin transient thermal imaging. Proceedings of the ASME 2010 
Summer Bioengineering Conference, SBC2010; June 15–19, 2010; 
Naples, Florida, USA.

 57. Pirtini Cetingul M, Cetingul HE, Herman C. Analysis of transient thermal 
images to distinguish melanoma from dysplastic nevi. Proceedings 
of the SPIE Medical Imaging Conference; February 12–17, 2011;  
Lake Buena Vista, FL, USA.

 58. Harris C, Stephens M. A combined corner and edge detector. 
Proceedings of the 4th Alvey Vision Conference; 1988.

 59. Hartley R, Zisserman A. Multiple View Geometry in Computer Vision. 
Cambridge: Cambridge University Press; 2003:32–33.

 60. Grady L. Random walks for image segmentation, pattern analysis and 
machine intelligence. IEEE Trans Pattern Anal Mach Intell. 2006; 
28(11):1768.

 61. Odobez JM, Bouthemy P. Robust multiresolution estimation of 
 parametric motion models. J Vis Commun Image Represent. 1995; 
6(4):348–365.

 62. Pawley JB. Handbook of Biological Confocal Microscopy. 2nd ed.  
New York: Plenum Press; 1995.

 63. Webb RH. Confocal optical microscopy. Rep Prog Phys. 1996;59(3): 
427–471.

 64. Rajadhyaksha M, Zavislan JM. Confocal reflectance microscopy 
of unstained tissue in vivo. Retin Lipid-Soluble Vitam Clin Pract. 
1998;14(1):26–30.

 65. Rajadhyaksha M, Grossman M, Esterowitz D, Webb RH, Anderson RR. 
In vivo confocal scanning laser microscopy of human skin: melanin 
provides a good contrast. J Invest Dermatol. 1995;104(6):946–952.

 66. Busam KJ, Hester K, Charles C, et al. Detection of clinically amelanotic 
malignant melanoma and assessment of its margins by in vivo confocal 
scanning laser microscopy. Arch Dermatol. 2001;137(7):923–929.

 67. Charles CA, Marghoob AA, Busam KJ, Clark-Loeser L, Halpern AC. 
Melanoma or pigmented basal cell carcinoma: a clinical-pathologic 
 correlation with dermoscopy, in vivo confocal scanning laser microscopy, 
and routine histology. Skin Res Technol. 2002;8(4):282–287.

 68. Marghoob AA, Charles C, Busam KJ, et al. In vivo confocal scanning 
laser microscopy of a series of congenital melanocytic nevi suggestive 
of having developed malignant melanoma. Arch Dermatol. 2005; 
141(11):1401–1412.

 69. Scope A, Benvenuto-Andrae C, Agero AL, et al. In vivo reflectance 
confocal microscopy imaging of melanocytic skin lesions: Consensus 
terminology glossary and illustrative images. J Am Acad Dermatol. 
2007;57(4):644–658.

 70. Aghassi D, Anderson RR, Gonzales S. Confocal laser microscopic 
imaging of actinic keratosis in vivo: a preliminary report. J Am Acad 
Dermatol. 2000;43(1):42–48.

 71. Busam KJ, Hester K, Charles C, et al. Detection of clinically amelanotic 
malignant melanoma and assessment of its margins by in vivo confocal 
scanning laser microscopy. Arch Dermatol. 2001;137(7):923–929.

submit your manuscript | www.dovepress.com

Dovepress 

Dovepress

Dovepress

212

Herman
 

C
lin

ic
a

l, 
C

o
sm

e
tic

 a
n

d
 I

n
ve

st
ig

a
tio

n
a

l D
e

rm
a

to
lo

g
y 

d
o

w
n

lo
a

d
e

d
 f

ro
m

 h
tt

p
s:

//
w

w
w

.d
o

ve
p

re
ss

.c
o

m
/ 

b
y 

1
2

8
.1

8
2

.8
1

.3
4

 o
n

 0
6

-A
p

r-
2

0
2

1
F

o
r 

p
e

rs
o

n
a

l u
se

 o
n

ly
.

Powered by TCPDF (www.tcpdf.org)

                               1 / 1

http://www.dovepress.com/clinical-cosmetic-and-investigational-dermatology-journal
http://www.dovepress.com/testimonials.php
www.dovepress.com
www.dovepress.com
www.dovepress.com
www.dovepress.com

	Publication Info 2: 
	Nimber of times reviewed: