PreprintPDF Available
Preprints and early-stage research may not have been peer reviewed yet.


Content may be subject to copyright.
Ramy Abdlaty1, Qiyin Fang1, 2*
1: School of Biomedical Engineering, McMaster University,
2: Engineering Physics, McMaster University,
1280 Main street West, Hamilton, Ontario, L8S 4K1
Erythema is superficial redness of the skin. Skin erythema may present due to many causes. One of the common
causes is prolonged exposure to sun rays. Other than sun exposure, skin erythema is an accompanying sign of
dermatological disease such as psoriasis and acne. Quantifying skin erythema, in patients, enables the dermatologist
to assess the patient’s skin health. Therefore, quantitative assessment of skin erythema was the target of several studies.
The clinical standard for erythema evaluation is visual assessment. However, the former standard has some
imperfections. For instance, it is subjective, and unqualified for precise color information exchange. To overcome
these shortcomings, the past three decades witnessed various studies that aimed to achieve erythema objective
assessment, such as diffuse reflectance spectroscopy (DRS), optical/ non-optical imaging methodologies, and
electrical measurement techniques. This current review article revises various studies in the past three decades and
discusses their methodologies, mathematical tactics for computation, and their limitations. In conclusion, the
limitations of the earlier erythema assessment techniques are motivation for developing novel techniques.
Keywords: Skin, Erythema, cutaneous diseases, akin inflammation
1 Skin Erythema
Skin erythema, or flare, is the reddening reaction of
the skin as a result of an external stimulus [1],
immunological reaction with/out hypersensitivity to
an allergen [2], or viral infection [3]. The flare size
depends on multiple parameters, for instance, the
distribution of the neural fibers and vascularization of
the stimulated region. Likewise, the strength and the
nature of the stimulus are factors that influence the
peak size of the flare. Typically, the flair’s peak
intensity is reached shortly after the stimulus onset. In
some cases, the flare is a result of an accumulative
process, such as radiotherapy treatment for cancer.
Bruce and Lewis interpreted the flare reaction to be an
axon reflex [4], [5]. Simply put, if a pernicious
provocation occurred to the human skin’s afferent
nerve, it would alert skin fibers. The fibers, in turn,
would respond to the alert by generating action
potentials mainly to the spinal cord, where they are
distributed to complementary axons. This leads to the
dispensation of neuropeptides like P /calcitonin gene-
related peptide (CGRP). The neuropeptides trigger
vasodilation, increasing blood to the skin, and thereby
creating a red appearance [1]. In a second
interpretation, erythema is the output of an
inflammatory cutaneous reaction associated with
diseases such as acne, psoriasis, melasma, as well as
fever. In severe skin reactions, skin erythema turns to
superficial blistering [6][8]. The blisters are
associated with wet desquamation that may take place
due to exposure to certain bands of electromagnetic
waves (EMW).
EMW, be it ultraviolet (UV), visible (VIS), or
infrared (IR) waves, are external stimulus of erythema.
Each of the mentioned spectral bands has an exposure
threshold or ‘minimal erythema dose’ (MED) [9]; if
reached, skin erythema is directly induced [10]. An
intriguing example of MED is the transient flush
erythema which takes place rapidly in people with fair
skin, in the summer upon exposure to sunlight [10]. In
addition to EMW exposure, physical pressure [11],
skin ulceration [12], [13], application of
cosmetic/medical topical agents, and electrical
stimulation [14][17] are all external stimuli of skin
erythema. Over and above, burns induce erythema
around resultant scars [18].
Radiation dermatitis [19][23] is typically an
equivalent term to radiotherapy-induced erythema. In
this case, erythema is a cancer radiotherapy treatment
linked side effect. The dermatitis reaction is
interpreted as a skin response to damage to basal cells
present in the epidermal layer. To ameliorate the
damaged region, deeper skin layers proliferate to
replace the impaired, superficial [24]. The radiation
dermatitis MED trigger is inconstant. However, there
are patients’ parameters, including skin type, age, and
tumor specifications, that affect dermatitis conduction
[25]. Despite the lack of knowledge on dose-intensity,
there is a rough threshold dose for triggering skin
erythema, above which causes skin reactions [25]. It is
of interest determine if permanent skin color change
occur in the case of intensity-modulated radiation
therapy (IMRT) due to the reduction of the
melanocytes in the irradiated region
[24],[22],[25],[29]. Dermatitis does not only indicate
underlying disease and compromises the patient’s
physical appearance, but also acts as an early warning
sensor for possible treatment pause [27]. Based on
prior facts, it becomes obvious that radiation
dermatitis calls for systematic monitoring and
quantitative assessment.
The purpose of this paper is to to review the techniques
utilized by earlier studies of skin erythema assessment.
The review is expected to produce a solid background
and foundation for developing innovative approaches
to quantitative and objective skin erythema
2 Assessment Techniques
A major goal for any skin erythema assessment
technique is to objectively quantify the redness
without the need for a skin biopsy or direct contact. A
potential approach is a contacless technique that
generates a real time graded redness intensity map.
Moreover, it is anticipated that the erythema
assessment standard device is miniaturized, easy to
operate, and costeffective. This section reviews the
techniques that were employed within the last three
decades , to evaluate, grade or detect skin erythema.
The review begins with visual assessment (VA) since
it is the gold standard for skin erythema evaluation.
Next to VA, common optical modalities used for skin
erythema assessment are introduced. Other, less
common erythema assessment techniques are
highlighted in section 2.3. Alongside erythema
assessment methodologies, various tactics of
computing the erythema index are overviewed.
2.1 Visual Assessment (VA)
VA is the original [28] and the simplest technique for
skin erythema assessment. VA’s simplicity is a result
of its tool/ equipment-independent nature. VA
depends solely on an experienced radiotherapist to
grade the intensities of skin erythema [29] based on a
predetermined ordinal scale of grades. Erythema
ordinal scales vary in the resolution between the
minimum and the maximum redness intensities. For
example, there are three [29] , four [30], [31], and ten
[28] steps erythema scales used in the literature. One
exemplary model, of the erythema scales, is shown in
Table 1 [28]. Although these scales are convenient for
rapid assessment, they lack linear transitions between
equidistant steps and thus they may miss or exaggerate
with distinct skin erythema intensities [32]. In addition
to the nonlinearity of erythema ordinal scales, VA
suffers from subjectivity, intra-observer variability
dependence, and incapability of precise color
communication. Despite all of VA’s disadvantaged, it
is still widely used compared to more objective,
engineered solutions. VA is (1) preferred because it is
fast, allowing for more patients to be seen in short
period of time; (2) simple to use, as there is no need
for operating/ adjusting any equipment; and (3)
suitable to the fast-paced clinical environment, since
the clinician can move easily between rooms [33].
Table 1: Example of an ordinal grading scale used for skin erythema
visual assessment divided into 10 different qualitative reactions [28]
Erythematous reaction
No reaction
Marginal reaction
Slight perceptible erythema
A greater than slight reaction which is not
sufficient to be classed as distinct
Distinct erythema
A greater than distinct reaction which is
insufficient to be classed as well developed
Well developed, possibly spreading erythema
A greater reaction which is not sufficient to be
classed as strong
Strong, deep erythema which may extend
beyond the treatment site
A more intense reaction than above
2.2 Colorimetry Assessment (CA)
Regarding human skin, color is a critical and
informative descriptor for both clinical and research
purposes [34][37]. However, the acquisition and
communication of color information visually are
limited by the time dependency, the describing
language, and the means of communication. For
instance, although the human eye is capable of
distinguishing between an enormous range of similar
colors [38], linguistic description is unable to
satisfactorily convey distinct color tones. The lack of
color descriptors to communicate various tissue color
tones in a quantitative manner motivated the
development of engineered colorimetry tools/
instruments to detect, analyze and archive distinct skin
color changes.
Color standards are well-known tools used to assess
the progression of some skin diseases associated with
skin color change. One of the earliest standards used
for this purpose was a set of red colored papers with
graded red intensities. A direct application of the red
papers set to the skin was used to demarcate the skin
erythema induced by UV exposure. The use of the
colored papers enabled the generation of UV dose
response curves [39]. The red set of papers were
neither durable, nor reliable; since it is easily worn and
hard to be typically reproduced. To overcome the
shortcomings of red papers, reliable and longer lasting
red photographic filters were developed as the
alternative. Unfortunately, neither red papers nor
photographic filters were sufficiently robust to
quantitatively grade skin erythema. Consequently,
both standards rapidly went out of use. As a result,
research interest transitioned to developing novel,
reliable, and inexpensive color scales.
Toward a convenient color scale, Taylor
hyperpigmentation scale (THS) was developed. THS
is based on an extensive statistical study of different
skin colors. The study employed a great number of
people of different racial backgrounds to obtain the
maximal skin color divergence [33]. Consequently,
THS developed 15 skin hue representative cards. Each
card is divided into ten graded pigmented steps. As
such, THS has a total of 150 different colors that are
expected to represent all skin types at a confidence of
95% [33], [40]. Meanwhile, a Japanese group
developed a new color tone scale based on Munsell
color system [41]. The Japanese color tone was printed
on flexible plastic bars. Each bar is divided into ten
color tones. The entire set of bars was used in a clinical
study to assess the laser treatment of solar lentigo [42].
The study was performed over approximately one year
and recruited 81 Japanese female patients [42]. Skin
color tone matching, by the aforementioned color
scales, was achieved in two steps as shown in Figure
1Error! Reference source not found.. The first step
was identifying the best hue that corresponds to the
participant’s skin color. The second step was scanning
the identified hue-card to find the best matching
hyperpigmented tone of the subject’s skin as shown in
Figure 1 [42].
The advantage of the color scales is the simplicity of
operation and the potential quantification of skin color
change which, in turn, facilitates easy data
communication and archiving. Although color scales,
or color-order systems [43] expanded the spectrum of
describable colors, it is still far from being objective
and vision acuity-independent. Once again, color
scales are susceptible to wear and tear, which renders
the color tones less discernible. Hence, there was a
need to depend on instrumental solutions to function
as robust colorimeters, in order to overcome
subjectivity and attain more reliable outcomes.
Figure 1: The process of allocating a color tone for human face
commences with finding the closest possible match through the
following steps, (1), a rough estimate is primarily determined by the
operator and then (2) more precise pigmentation color is selected
within left, right move, (3) followed by up and down adjustments.
In order to overcome the limitations of color charts,
systemic colorimeter devices were in demand. The
colorimeter device function is to identify the tissue’s
apparent color using a 3-dimensional color space. The
most common color space is the standard Commission
Internationale de l’Eclairage (CIELab), which is
represented by three axes: L*, a*, and b* [44]. The
first axis is the gray-scale, symbolized by L*. The gray
axis is divided into 100 divisions; starting from 0,
(complete darkness), and ending at 100 (bright white).
The second axis in CIELab space is the red-green axis,
symboledized by a*. a*-axis has the red intensity on
the positive side and the green intensity on the
negative side. The third axis is the yellow-blue axis,
symbolized by b*. b*-axis has the yellow intensity on
the positive side and the blue intensity on the negative
side. Both a* and b* axes are, in principle, applications
of Hering’s opponent color theory [45], which
explains the human vision sensistivity to color
detection. In the CIELab color space, the hue angle
is defined as the relevant psychometric of the
visually perceived property of hue such as red, green,
and magenta, and it is calculated using equation (1). It
is prudent to know that the perceived color is not pure
hue but is influenced by the color saturation as well.
The color saturation is expressed by the displacement
on the L* axis and termed chroma C, in the CIE.
Chroma is calculated by equation (2) .
 
  
Color detection accuracy, in CIELab, is dependent on
three factors, 1) the illuminating light specifications,
2) the light modulation by the tissue under test, and 3)
the human vision attributes. CIELab strictly delineates
these factors. First, the CIELab specifies the standard
illumination for color measurements as published in
tabulated formats [44], [46], [47]. Second, CIELab
identifies the standard for the spectrometer applied to
measure the light modulation in the range of the VIS
region [46]. Third, the human visual system was
investigated and the three matching color functions: ,
, and
were determined. There are three matching
colour functions because human vision interprets color
using a trichromatic system [47][49]. The CIELab
color system is a convenient, robust way for
communicating color information; moreover, it is able
to measure the differences between the perceived
tissue colors. The difference (∆E) between two colors
is represented by the square root of the displacement
of both colors in L*a*b* coordinates as shown in
equation (3) [44].
  
2.3 Spectra-based Assessment
Spectral reflectance concept has been involved in the
objective assessment of the skin color for a long time
[40]. This concept has contributed in developing
robust and reliable instruments such as the tristimulus
instruments and the narrow-band spectrometers. The
tristimulus instruments, such as Minolta chromameter,
depends mainly on the measurement of the reflected
light within three central wavelengths. Narrow-band
spectrometers, unlike tristimulus instruments, such as
the Dermaspectrometer, erythema meter, and
Mexameter, measure the reflected light in specific-
bands, mainly the red and green ones.
Tristimulus Colorimetry instruments were originally
developed to objectively and reliably measure color
information akin to the human visual system. Simply,
the method of measurement depends on illuminating
the tissue under investigation with a polychromatic
light source and detecting the back-reflected intensity
through three separate spectral channels. One or two-
dimensional array of photodiodes are used for light
detection. Minolta chromameter is a common example
of the tristimulus colorimeters [18], [50][55].
Minolta instrument illuminates the object using a
xenon light source and collects the diffusely back-
reflected intensity at three central wavelength bands:
450, 560, and 600 nm. The instrument uses a mapping
transformation to a color space such as CIELab
standard in order to elucidate the object color. Minolta
chromameter gains a good reputation due to its
robustness and replicability [53]. Based on this
reputation, Minolta chromameter color data was
examined in comparison to visual grading for
investigating skin blanching onset in response to the
application of a topical corticosteroid agent on
volunteers for three subsequent days [51]. The results
of the study verified the higher performance of
Minolta chromameter in distinguishing the blanching
effect as shown in Figure 2. The study, also, prooved
that the L* axis value is the most sensitive component
in the color space for discriminating the skin
Figure 2: (a) Algebraic sum (mean ± std) for the visual assessment
score, (b) the gray scale values , (c) the red-green values
, and (d) the yellow-blue values  for the skin color
change induced by applying a topical corticosteroid agent using the
Minolta Chromameter. [51]
The chromameter is inexpensive way for objectively
measuring skin color. However, the shortcoming of
such way is the necessity for systemic calibration each
time they are used. Hence, an overhead time is spent
to set up the instrument for operation. Therefore, it
became inappropriate to use chromameter in the daily
clinical practice.
Narrowband Spectroscopic Devices principle of
operation is based on the fact that the green and red
portions of the visible spectra suffice to estimate the
change in skin’s pigmentations: hemoglobin, and
melanin, in a quantitative approach [56][60]. For
instance, the skin redness/ erythema severity is
quantitatively estimated by subtracting the melanin
effect from the green filter absorbance [56]. Other
studies quantitatively evaluated erythema by only
computing the changes in the hemoglobin content and
negligibly consider the melanin’s effect [61], [62].
As a direct application, a study reported the
development of a portable device called the erythema
meter [63], [64]. The device measures the skin’s back-
reflected light within the red and the green spectral
bands. The measured back reflected light gives the
input to roughly estimate the cutaneous hemoglobin
content. The hemoglobin content estimation is based
on calculating the erythema index (EI), as displayed in
equation (4).
 
In conjunction with the erythema meter, two more
instruments are common in the class of the narrow
band devices: the Dermaspectrometer [56], and the
Mexameter [65]. The Dermaspectrometer device
integrates two light-emitting diodes (LED) to
illuminate the targeted skin ROI [56]. One LED emits
at 568 nm (green band) and the other emits at 655 nm
(red band). The skin’s reflected intensities at the two
wavelengths are detected and analyzed to provide both
the melanin and the erythema indices.
The Mexameter, unlike the Dermaspectrometer, uses
16 integrated LED sources combined in one probe
emitting at three separate bands; green (568 nm), red
(660 nm), and near-infrared (NIR) (880 nm) bands.
The Mexameter computes the skin’s melanin index
(MI) based on the measured red and NIR back
reflected light as demonstrated in equation Error!
Reference source not found.). The erythema index is
computed based on the measured back reflected light
in the green and red bands as demonstrated equation
Error! Reference source not found.) [60], [66]. A
sample type of the Mexameter is shown in Figure 3.
  
  
 
The Mexameter is more sophisticated and expensive
in contrast with the Dermaspectrometer. In terms of
advantages, it has an accuracy range of ± 5% provided
that three readings are obtained [65]. As a result, it is
commonly used in cosmetic and pharmaceutical
cutaneous research studies. Recently, it was approved
to be put to use in: skin melasma investigation [65],
skin toxicity diagnosis, and in monitoring melanin
content variation due to radiotherapy [67].
Figure 3: (a) the Mexameter device is equipped with a circular probe
of 5 mm diameter and 20 mm2 surface area. (b) Example of using
the Mexameter to quantitatively assess the skin damage attributable
to breast cancer radiotherapy treatment [67].
The narrow-band spectroscopy principle is an active
approach for objective assessment of skin erythema.
Thus, it has been used in a real-time measurement for
an artificial induced skin erythema study [11]. The
study used an RGB sensor to measure the skin
erythema due to an instantenous applied pressure, and
monitor the temporal evolution of erythema in case of
a constant pressure. In sum, the narrow-band devices
are able to provide successful figures for the two skin
pigmentation indices: the erythema and the melanin.
However, they lack the potential for fully interpreting
the skin color changes. Hence, the measurement of the
skin’s total diffuse reflectance was a more appropriate
approach to perform this interpretation.
Diffuse reflectance spectroscopy (DRS) systems are
involved in the measurement of the total diffusely
back-reflected polychromatic light out of an
illuminated object. Regarding human skin, the total
diffusely reflected light is the summation of the back-
reflected light excluding the specular reflected light.
The excluded portion of the reflected light accounts
for only 4:7 % of the total reflected light [68].
However, it probes the skin surface information
including the refractive index. Conversely, the
diffusely back-reflected light checks out the structures
beneath the tissue surface, such as the underlying
vascularization and pigments, by scattering inside the
tissue before bouncing back. DRS measurements can
be accomplished using both time domain- [69], [70],
and frequency domain- [71], [72] based techniques.
DRS has been used in investigating divergent skin
signs associated with cutaneous diseases [21], [73]
[75]. DRS systems were implemented utilizing
multiple optical configurations including the
integrating sphere (IS)-based DRS [76][79] and the
spatially resolved steady-state DRS [80][82]. To get
more in depth, the IS and the spatially resolved based
DRS techniques use continuous-wave (CW)
illumination sources and a spectrophotometer for
measuring the back-reflected spectra. DRS systems
are advantageous in measuring the tissue’s optical
properties, due to their simplicity, compactness, and
low cost.
For instance, a recent study proved that IS-DRS
system can be built with low costs while being
efficient and simple to construct as shown in Figure 4
[17]. The same setup was used in a clinical study to
detect the temporal development of skin erythema via
computing the erythema index. Figure 5 displays the
computed erythema indices in the study, based on the
measurement of the daily skin diffuse reflectances in
cancer patients. Based on the computed erythema
index, DRS was able to detect the patients’ skin color
change earlier than visual assessment, however, both
techniques were done synchronously [17]. Over and
above, the acquired DRS data were able to quantify
skin erythema via estimating the apparent
concentrations of skin chromophores during radiation
treatment [17]. In sum, the simple and affordable IS-
DRS system succeeded to properly detect and
precisely quantify radiation dermatitis during cancer
treatment before visual detection by radiotherapists
was possible. Unfortunately, DRS measurement
techniques can only be used for relatively small ROI
inspections with a low demand for spatial resolution.
Moreover, it requires direct skin contact, which in
some cases (dry and wet desquamation, burns) is hard
to achieve.
Figure 4: An inexpensive integrating sphere-based diffuse
reflectance spectroscopy (DRS) configuration used for skin
erythema measurement. The system is composed of a simple wide
band light source coupled to an integrating sphere which connected
to a PC controlled spectrometer [17].
Unlike both IS and spatially resolved DRS techniques,
the time domain (TD) based techniques utilize pulsed
light for illumination and detect the skin’s back
reflection on a temporal basis. Moreover, in order to
work properly, the source-detector displacement in
TD-DRS techniques should not exceed a couple of
centimeters range [83]. To put it simply, the TD- DRS
based techniques depend on sending a light pulse to
the tissue and detect the back reflected return. Due to
the tissue’s optical properties, the reflected light pulse
is modified in shape and attenuated in amplitude [84].
Based on the detected light pulse envelope
modification and the amplitude decrement, the tissue’s
optical properties is estimated.
Contrasting TD-DRS, the corresponding frequency
domain (FD) techniques of diffuse reflectance
measurement utilize an amplitude-modulated light
source and detect the phase and the amplitude
modifications at one or more ROI sites. Reduced
sensitivity to noise and straightforward optical
implementation are the strong advantages of FD-DRS.
Moreover, FD-DRS systems’ reliability in data
analysis is significantly larger than the corresponding
TD-DRS techniques [85].
Figure 5: An illustration of the computed melanin-corrected skin
erythema index on a daily basis during radiotherapy. The vertical
black line at the 18th day indicates the first day for the skin erythema
to be clinically observed by visual assessment [17].
Laser Doppler Flowmetery (LDF) is also known as
laser Doppler velocimetry. LDF is an optical
technique which enables measuring the speed of a
moving fluid. The fluid speed is computed via
detecting the induced frequency/ Doppler shift in a
crossing laser beam. In the context of skin erythema,
the moving red blood cells inside the superficial skin
vessels cause a Doppler shift, by which erythema
could be quantified. Since the erythematous response
was originally interpreted being dependent on
vasodilation of microcirculation [15], [86], [87].
Although LDF technique looks harmonious for
erythema quantification, a past study reported its
lower sensitivity in evaluating the UVB-induced
erythema. The study supported its report by comparing
the Minolta colorimeter and the spectroradiometer
[38]. It explained that the blood perfusion’s relative
change measured by the LDF is not as well correlated
with the skin color change as it is in diffuse reflectance
spectra [38].
2.4 Imaging-Based Assessment
Imaging was explored as an alternative to overcome
the limitations of the spectra-based measurement
techniques, including direct skin contact, limited ROI
investigation, and local region possible misguidance.
Since imaging-based techniques are contact-free, and
applicable to examining larger ROI. For instance, LDF
were substitited partially in novel applications of laser
Doppler imagers (LDI’s). LDI’s are equipped with a
CCD camera, through which false color images for the
region of interest can be produced. Thus, the induced
minimum erythema discernible (MED) becomes
easier to visualize, as shown in Figure 6 [66]. Despite
the advanced display in LDI, it is still limited to a small
area of investigation. In addition, it lacks the precision
of erythema quantification. To increase the area of
investigation, digital imaging was the gate.
Figure 6: (a) Laser Doppler imager (LDI) captures images for
human skin after applying two different doses of UVB irradiation,
and erythema was induced. (b) the corresponding color image of the
induced erythema [66].
Digital Color Imaging and Photography is a well-
suited approach to surmount visual assessment inter/
intra-observation particularity. It also beats the
limitation of the memory of the dermatologist who
struggles to retrieve the condition/s of the patients in
former sessions. In burns, digital photography spares
sensitive skin regions from any contact-analysis
during clinical work or research studies [88]. Taking
all together, the former advantages, of digital imaging,
provides a good opportunity for conducting more
research on lately emerging topics. This opportunity is
supported by the current technological breakthroughs
in electronics, communication, and photography at
low cost. For instance, digital imaging, amongst
others, have progressive capabilities including (1)
autonomous color correction, (2) huge data storage,
(3) wireless transfer of images via Bluetooth, and
WiFi, and (4) autofocusing. Similar to hardware
development, innovations in image analysis/
processing engendered significant steps in
registration, feature extraction, and classification.
As such, digital imaging becomes a low-cost portable
solution not only for skin pigments quantification
[89][93], but also in diagnosis of other cutaneous
diseases [94]. In support of high dynamic imaging,
four factors need to be attentively considered: (1)
illumination, it is necessary to be relentless in intensity
and spectra; (2) optimized selection of patient poses,
as the skin is highly scattering tissue [90]; (3) the use
of auxiliary tools such as special lenses, strong
flashlights, spectral band filters, and color checkers,
provide the optimal conditions for imaging; and (4)
intellectual image analysis algorithms need to be
developed, in order to retrieve embedded information
inside the captured images. Over and above, individual
skills and intermittent training for the photographer
play a serious role in producing informative photos. To
enhance the image quality, modern technologies
provide highly precise and automated photographic
apparatus like the one shown in Figure 7, nonetheless
they are eminently expensive.
Figure 7: An example of a high-end facial skin imaging station
equipped with multiple illumination styles and a ready-made skin
analysis software tool used for facial skin inspection [91],[95]. The
equipment is immensely expensive due to its independence on
human skills in imaging.
Regarding skin erythema imaging, Sergio Coelho et al.
[96] reported the feasibility of using digital
photography to quantify skin erythema induced by
long-term UV exposure. The study employed a digital
color image analysis algorithm named “CADIE” to
evaluate the skin’s erythematous response in
volunteers of different backgrounds based on CIELab
color system as shown in Figure 8. The study was able
to monitor the severity of skin erythema induced by
excessive exposure to multiple units of UV radiation.
The UV radiation unit is defined as the dose linked to
MED. However, MED is an property unique to an
individual’s skin type and physiology. Albeit all the
challenges faced the study, the use of CADIE
algorithm enables better identification of skin color
changes in terms of ∆a* relevant to the received MED
units. In spite of this efficacious work, it is important
to note that photography loses clarity in revealing
tissue surface/subsurface details, wheras polarization
does not. Accordingly, polarized light imaging was the
key solution to retrieve the lost lucidity.
Figure 8: An illustration of the erythematous response in distinct
backgrounds (W for white, A for Asian, B for African American)
volunteers. The top row, color photos of ROI in one subject from
each group of the formerly mentioned backgrounds. The middle
row, the relationships detected between the dose of UV radiation
expressed in MED units (MED unit is the minimum required UV
dose to induce a discernible erythema on the skin of a volunteer) and
CADIE-assessed red-green axis ∆a* after 1 day of UV exposure.
The bottom row, the same relationships are computed based on DRS
system using a spectrophotometer. [96]
Polarized light imaging provides the opportunity to
selectively acquire information from the tissue’s ROI
surface and the underneath layers [97], [98]. The
former opportunity is based on the fact that the light
reflected back out of any tissue’s surface, in the case
of linearly polarized illumination, is composed of two
components. The first of which, specular reflection,
probes for the tissue’s surface and texture. This
component retains the clone polarization type as the
incident illumination. In contrast, the second reflected
component, diffusely reflected light, does not conduct
the same illumination polarization. The loss of
polarization is accredited to the scattering interaction
of photons beneath the tissue surface, before the
photons bounce back. Therefore, the two components
reflected from tissue can be polarly distinguished. To
do that, a linear polarizer needs to be installed and
selectively oriented ahead of the imaging detector. For
surface information acquisition, the linear polarizer
should be matched to the incident illumination
polarization direction. To exclude surface data,
orthogonal polarization to illumination’s one has to be
chosen. In an erythema framework, polarized imaging
improved the skin redness visualization in acne lesions
[89]. Furthermore, improved contrast was reported as
a benefit of polarized imaging in monitoring the
microcirculation, vasodilation, and vasoconstriction as
shown in Figure 9.
Inspired by the considerable improvement achieved by
polarized imaging, a recent study claimed that linear,
elliptically polarized light imaging is also a potential
approach for enhancing the contrast and the resolution
depth of captured images [99]. Although polarized
imaging did a great job in visualizing particular
polarization oriented tissue features, other tissue
symptoms still can not be elucidated except if excited
earlier, as in fluorescence imaging.
Fluorescence imaging detects prompted emission,
and mainly fluorescence, as it may be more valuable
than absorption property in scaling molecules [100].
Fluorescence imaging procedure depends on a limited
spectral band for both excitation, and detection.
Filters/ monochromators are always incorporated in
the optical path to ensure only excitation and emission
bands are transmitted. Fluorescence measurements of
the human skin are frequently done using optical fibers
since excitation and emission bands are easily
separated [37]. In general, the involved fiber bundle,
in measurement, is divided into two pathways: one to
deliver the excitation light while the other pathway is
to collect the skin emission.
Figure 9: The top row, the normal skin image, for a bounded region,
is displayed captured with digital photography on the left and with
spectroscopy details on the right. The bottom row shows the same
region of interest images, by both photography and detailed
spectroscopy, however, showing the effects of using a vasodilator
drug, induced by iontophoresis. Polarization spectroscopy imaging
proved to be more efficient in displaying the heterogeneity of the
microvasculature over the whole studied region of skin [101]
Considering skin as a sample, the chromophore,
melanin, significantly absorbs light in the wavelength
range 340:400 nm and absorbs relatively less light in
the emission bands 360:560 nm [102]. The optical
properties of melanin provide fluorescence imaging
with a good likelihood to identify the highly
concentrated regions of the chromophore distribution
within the skin. In the context of erythema,
fluorescence imaging was used in a couple of studies,
for examining post-inflammatory hyperpigmentation
(PIH) related to acne lesions and monitoring
superficial acne spot development due to courses of
treatment [103], [104].
A simple schematic diagram for fluorescence imaging
optical configuration is shown in the top row of . The
figure illustrates the basic setup of fluorescence
imaging. A wideband light source is (xenon in this
case) filtered to excite the tissue under investigation.
The prompted emission of the excited tissue is filtered
by an emission filter and then imaged using a camera
In Figure 10, the middle row shows two photographic
pictures for a patient who has facial acne lesions
before and after 12weeks of treatment. The bottom
row displays the same patient’s fluorescence images at
the formerly mentioned time points. By image
analysis, the study, to which the images belong to,
proved that fluorescence images of the patient were
more capable of highlighting the facial erythematous
regions’ modified due to the used treatment in contrast
with regular photography [104]. Even if fluorescence
imaging is required to show molecular scale particles
in biological tissue, it is neither the safest approach nor
the only way to achieve such information. For
instance, microscopic imaging is also intertwined with
erythema assessment in order to create a deep
understanding of the linked tissue microstructure.
Figure 10: Top row, a schematic diagram of a fluorescence optical
imaging configuration. Middle row, an acne patient’s images with
flash photography is displayed at week-1 (left) and week-12 (right)
of acne treatment. Bottom row, the images of the same patient, at
the same weeks of treatment, however, captured by fluorescence
imaging configuration are shown. The fluorescence images were
delineating the acne lesions in the face more clearly than traditional
photography [104].
Microscopic imaging is an approach to exploring skin
diseases associated with erythematous signs at the
cellular scale of resolution. Microscopic imaging
enhances the assessment accuracy of the skin’s
inflammatory regions by permitting the visualization
of the detailed structures encompassed in the skin
layers. For instance, the stratum corneum and
dermoepidermal junction is difficult to visualize using
macroscopic imaging, and instead requires
microscopy. It is interesting how much details can be
revealed with a microscope like Dermoscopy. For
example, Dermoscopy is capable of disclosing three
distinct skin properties: architecture, pigmentation,
and divergent regions’ perimeters. As a consequence,
it provides a great opportunity for physicians to
interpret rare dermatological features [105].
Confocal microscopy is a distinguishable functioning
microscopic imaging configuration [106]. It is capable
of resolving microanatomic structures within the skin
to a near histological resolution [107]. For instance,
confocal scanning laser microscopy brings to the
dermatologist's eyes detailed skin structure without the
need for biopsy. This scale of resolution is very
explnatory in skin diseases including contact
dermatitis, as shown in Figure 11 [108], [109].
Figure 11: Sections of skin diagnosed with allergic contact
dermatitis (ACD) and other sections for control healthy skin regions,
where the confocal microscopy images show the differences in the
skin structure between ACD skin regions and the normal ones at a
near-histological level of resolution [108], [109].
Confocal microscopy is comparable both histology
and patch testing, in diagnosing and monitoring
allergic contact dermatitis (ACD), due to its high
sensitivity and specificity in detection [110]. Although
dermatological microscopy offers high resolution, it
has some disfavors. The disfavors of microscopy are:
(1) the instruments are bulky, which makes it difficult
to transport; (2) pricey, thus it is not affordable in
small dermatological clinics; and (3) most of the time
invasive, thus inconvenient for patients; and (4) time-
A less expensive, yet still highly informative
technique is spectral imaging. Spectral imaging is an
emerging technique because of its capability of
acquiring a wealth of information in both the spatial
and the spectral domains.
Spectral Imaging (SI) is an imaging technique in
which multiple frames are taken for the object of
interest, like in photography, however, at distinct
wavelengths. Thus, the total spectral radiance of the
object is acquired. SI generates a 3-dimensional data
set, one dimension is spectral and the rest is spatial.
Different names are given to SI such as imaging
spectrometry/ chemical imaging [111], [112]. A list of
parameters are utilized as basis for SI classification
including the number of bands, resolution, and
acquisition schemes.
Data acquisition in SI is accomplished via different
ways including scanning (spectral or spatial) and non-
scanning (snapshot) techniques. To give an example,
the spectral scanning technique is dependent on
involving optical filters which might be static or
tunable filters. The tunable filters (TF) scan the spectra
with high resolution and does not produce vibration in
a very short time. These advantages allow the TF-SI to
dominate in SI configurations. Two famous examples
of tunable filters are; acousto-optic (AOTF) or liquid
crystal (LCTF). SI, in general, imposes the use of
broadband light sources that produce a broadband of
emission, such as tungsten-halogen or xenon lamps.
This emission is sliced and detected by a
monochromatic CCD or CMOS sensor.
Regarding SI number of bands, for example, there are
three classes; multispectral imaging (MSI, usually
uses less than 10 bands) [12], [113], [114],
hyperspectral imaging (HSI, > 10 and < 1000 bands)
[115][117], or ultraspectral imaging (USI, ≥ 1000
bands) [118]. For the sake of convergence, only HSI
will be discussed.
HSI produces a datacube that holds a wealth of
information. This wealth of information offers a
potential solution to arduous obstacles in a sundry of
applications [119]. To give an example, HSI was used
to address many challenges in the medical field as
reviewed elsewhere [120]. Particularly, HSI is a
technique of great potential for skin erythema
assessment because it is contactless, objective,
suitable for wide area imaging, and capable of
mapping chromophore distributions. Taking all these
reasons into account, HSI is likely an excellent
candidate for skin erythema quantitative assessment.
The challenge of objectively quantifying skin
erythema that is non-homogenously distributed over a
wide skin region can likely be resolved with a
technique which combines the advantages of imaging
and spectroscopy in one equipment. This equipment
should perform hyperspectral imaging and produce
spectral-spatial datacube. The intellectual analysis of
the produced datacube enables mapping of the skin
color changes relevant to the induced erythema, as
wekl as calculating the apparent concentrations of the
skin chromophores [121] [122]. HSI has been used in
literature to delineate the variable intensities of skin
erythema accompanied by a few skin problems [122]
[126]. For example, HSI was capable of mapping acne
lesions, the skin’s viral infections, and contact allergic
dermatitis, which are associated with erythematous
signs as well as erythema temporal evolution [127].
Morever, HSI is more proficient at mapping the
erythematous skin regions in terms of contrast,
compared to red-green-blue (RGB) digital imaging.
A schematic diagram and implementation of a recently
developed multiview HSI [128] used for studying the
tissue structure and its functional attributes is shown
in Figure 12. The novel multiview capability of HSI
enabled the operator to outline the topography of a
wound associated with erythematous regions. In
addition, it facilitates the construction of a 3-
dimensional image as shown for a wound model in
Figure 13 [128].
Figure 12: Multiview HSI schematic diagram, laboratory
implementation. The system is composed of two cameras; one is
digital and the other is spectral, where both cameras are moving
along a horizontal arm for positioning and the horizontal arm is
leveled vertically with a step resolution of 0.5 mm. A mirror set is
integrated between the object and the spectral camera to enable
Multiview process [129].
Several studies ([93], [130], [131]) put an effort to
obtain the erythema maps from RGB images;
however, the quality of these maps was very limited
[132]. The former limitation was attributed to the
skin’s optical properties which are spectrally
dependent on the encompassed chromophores (both
types of hemoglobin, melanin, water,…). Therefore, it
becomes difficult to optically monitor and precisely
quantify the chromophores concentration alterations
using only red, green, and blue channels in
photography. As a result, HSI is tagged the “gold
standard” for skin erythema mapping [132]. Added to
mapping, HSI has an impressive sensitivity to detect
the onset of erythema before it becomes visible to the
clinician eye [132].
Figure 13: Top-row, (a): Multiview-HSI system (b, c) the foot
wound model, (d) digital color pictures taken by the Multiview
modality. Bottom-row, (a) 3-D wound model topology built from
the Multiview hyperspectral captured frames, (b) and the
corresponding digital color 3-D wound model. [128]
3 Other Assessment Techniques
This section is dedicated to presenting an overview of
the less frequently utilized techniques for skin
erythema assessment including optical coherence
tomography, ultrasound imaging, magnetic resonance
imaging, and dielectric constant measurement. The
reasons for the low popularity of these techniques are
diverse. For instance, these techniques are disfavored
by lower sensitivity, higher cost, complexity,
substantial size and weight, mobility problems, or a
combination of the above.
3.1 Optical Coherence
Tomography (OCT)
OCT is an optical technique that makes use of the
interference of coherent light. To put it simply, OCT
imaging uses a Michelson interferometer, as shown in
Figure 14 top-row [133]. The information in OCT
stems from the interpretation of the interference
pattern generated by a low coherence laser beam. The
laser beam in OCT configuration is split into two rays,
one of which is directed to the sample and the other is
guided toward a mirror. In-phase interference
generates bright fringes, constructive interference, and
dark fringes, destructive interference. The laser’s short
coherence length permits the determination of the
penetration depth: images from subsurface layers can
be drawn in analogy to A-scans in ultrasound [37].
While B-scans in OCT are formed by the combination
of fringe intensities in adjacent A-scans.
In the past, OCT required a long time to build a tissue
image due to its slow scanning techniques. Real-time
imaging is now possible [134]. The penetration depth
of OCT is mainly dependent on the laser’s central
wavelength. For example, the NIR wavelength of 1300
nm leads to a penetration of about 1.2 mm in
comparison with 0.7 mm using light in the 700 nm red
region. The typical lateral resolution may vary
between 10-15µm, however, there are some special
techniques that can lead to 1-3µm penetration,
entering the region accessible by confocal microscopy
[133]. OCT has the feasibility to detect the presence of
skin diseases associated with tenderness and erythema.
OCT’s detection of erythema is observed on the
interference built images in the format of increased
thickness in the epidermal layer, reduction in light
scattering, and a dilated vasculature network [135], as
shown in Fig.14 bottom-row [134]. Although some
OCT systems are currently available as bench-top
devices, their sensitivity is utterly vulnerable to
change in the case of slight movements during in vivo
measurements. These unintentional movements of
patients result in blurry images [134]. Two major
differences exist between OCT and ultrasound: need
for skin contact and the sensitivity to patient
Figure 14: (top) Schematic diagram of a high definition OCT
(reproduced from [133]), (bottom) 2D-OCT scanned images of
human back skin before (left) and after (right) UVB irradiation;
after radiation, the stratum corneum is separated from the
subsurface layers and the light intensity is reduced [134]
3.2 Ultrasound-Based Imaging
Ultrasound, like OCT, is supplementary equipment in
the dermatological bench-top instrumentation.
Ultrasound became a tool to measure the skin
thickness since the last quarter of the 20th century
[136]. Following that, ultrasound imaging expanded to
different dermatological issues, including skin
irritation and erythema [137][141]. The basic
principle of ultrasonic imaging is detecting the
reflected acoustic waves exiting from the different
layers within the tissue of interest. The tissue
components are quite often different in density and
hence they respond to the transmitted acoustic waves
individually. Distinct responses, from the different
layers of the scanned tissue to the acoustic waves, are
transformed and displayed on the ultrasound monitor
as an image of the tissue’s underlying structure
represented by gray scale intensities [142]. A wide
range of acoustic frequencies is used in ultrasonic
imaging, where lower frequencies are used to visualize
deeper structures in tissue. Examples of the used
frequencies are shown in Table 2 [142]. In advanced
ultrasonic imaging, the system is not only imaging
stationary structures but also monitoring the moving
fluids like arterial/ venous blood by adopting Doppler
ultrasound methodology.
Table 2: Ultrasonic imaging frequencies and associated penetration
depth within visualized skin part [142]
freq. (MHz)
Visual skin part
depth (cm)
Lymph nodes
~ > 4.00
Dermis and epidermis
~ 3.0:0.3
Dermis and epidermis
~ 0.6:0.7
~ 0.3:0.015
Ultrasonic imager is a noninvasive and portable
device, used to measure the tissue thickness within the
lesions area. The tissue thickness was found to be
inversely proportional with skin erythema. This
relationship between the skin thickness and the skin
erythema due to irradiation was verified in a breast
cancer study [137]. Another study showed the
productive usage of ultrasound in imaging a forearm
dermatofibroma nodule surrounded by an
erythematous region. The nodule was imaged, blue
dot, before the removal surgery (see Figure 15 [143])
to alert the surgeon of the nearby vein. Although
ultrasound was able to quantify erythema indirectly
via skin thickness measurement, a contradicting study
showed that an acoustic equipment was less
efficacious at assessing erythema [67]. The
contradicting study measured skin thickness using an
ultrasound-based device, and showed a low correlation
between both the melanin and erythema indices.
Hence, the study concluded that skin thickness is not a
productive measure of assessing subcutaneous fibrosis
and the componential changes of the skin parenchyma
Therefore, ultrasound is an inadequate technique for
studying skin erythema associated with different
cutaneous diseases in contrast with, for example,
spectral based modalities.
Figure 15: An illustration of Dermatofibroma (DF) case inspected
by ultrasound instrument: (A) forearm erythematous nodule, (B)
Ultrasonic transverse view image displays erythematous nodule
near a large vein (blue spot) which assist in being avoided in
surgery. [143].
3.3 Magnetic Resonance Imaging
MRI is a distinguishable imaging technique by which
high spatial resolution images for minute tissue
components can be acquired [144]. Such highly
sophisticated imaging modality is not commonly used
for superficial skin disorders. However, in a previous
study, MRI was used toward 3-D mapping of acne
lesions and psoriasis as shown in Figure 16 [145]. The
major disadvantages of such high-resolution imaging
modality are; the big size, the technical complexity,
and the instrument high costs.
Figure 16: (a) Histological image of a severe psoriasis case, (b)
anatomical view with the marked region of interest (ROI), (c) a
relaxation time T1 map calculated for the ROI, where the increased
red color intensity in the outer dermis layer is due to the skin
inflammation [145] .
3.4 Dielectric Constant
Dielectric constant measurement is used in diverse
medical applications as reviewed elsewhere [146].
Regarding the skin, dielectric constant measurement
was utilized to detect the reactions induced by
irradiation in breast cancer radiotherapy treatment
[147]. The basic way to measure the dielectric constant
of the skin is to connect an open terminal coaxial probe
to a network analyzer. The probe functions to transmit
electromagnetic (EM) waves at harmless frequencies
through the skin’s ROI and to measure the return
signal through the network analyzer. It has been
reported that the dielectric constant is inversely
correlated to the clinically scored erythema [147].
Hence, dielectric constant measurement proved to be
an objective method of noninvasively quantifying skin
erythema. Nevertheless, the skin variability between
people and the inconsistent dielectric properties in the
same person yield less repeatable results. Therefore
dielectric constant measurement was less credible for
clinicians to use in their daily practice.
4 Erythema Indices
This section briefly introduces the different
computation approaches for the skin erythema index
through a review of the literature. The purpose of this
review is to hand over a synopsis of the skin erythema
index calculation in order to encourage more studies
to validate and compare them [148]. This comparison
helps to identify the optimal skin erythema estimation
4.1 Dawson Indices
The use of Dawson indices is a familiar approach for
quantitative estimation of the skin color changes due
to a disease’s development [149], or a medical
treatment [68]. Dawson derived his erythema index
(DEI) formula based on the assumption of a
proportional relation between the hemoglobin optical
absorption in the range of 510-610 nm and the
erythema scores. Dawson found that the area under the
curve for the hemoglobin absorption, if a baseline is
determined, is the parameter related to the skin
erythema variation and calculated as follows:
   
The terms, p, q, r, s, and t are the symbols expressing
the logarithm of the reciprocal of the reflectance
(LRR) measured at five predetermined wavelengths
510, 540, 560, 580, and 610 nm, respectively. Dawson
melanin index (DMI) is computed based on two
central wavelengths (650 and 700 nm). As it is
assumed that melanin concentration is relative to the
slope of the skin spectral curve of absorbance (
) bounded by the two formerly mentioned
wavelengths. The formula used for computing melanin
index is the average of the skin absorbance values for
two wavelengths surrounding 650 nm and 700 nm
 as seen in equation 8:
   
 
The term  is the computed reflectance at certain
wavelength in nanometers. The constant at the end
is an empirical correction constant. The DEI can be
corrected (DEIc) using the DMI by applying the
formula as shownin equation (9), where γ = 0.04 is an
empirically derived balancing constant by Dawson to
avoid negative values due to the existing while being
insignificant skin chromophores:
  
Subsequent reflectance measuresments are taken to
evaluate response to pharmaceutical treatment of a
cutaneous disease. Following the masurements,
Dawson’s relative erythema index () is computed
to assess the skin color change. The term  is
basically equivalent to the difference in value between
 for two skin sites, one of which is the treated site
while the other is the control (non-treated) site.
4.2 Diffey Index
Diffey index [62] was built on a simple observation of
hemoglobin absorption attributes. Diffey realized that
hemoglobin absorption is higher in the green spectral
region than the corresponding red region. Based on
this realization, Diffey proposed an erythema index by
computing the ratio of the inverse logarithm of the skin
spectral reflectance at two wavelengths: green (565
nm) and red (635 nm). Diffey didn’t correct for the
melanin effect in his computation. The formula which
expresses Diffey index is given in equation 10.
  
4.3 Tronnier Index
Tronnier [150] had a similar approach to erythema
computation as Diffey [151]. A difference is that
Tronnier principle computes the relative rather than
the absolute erythema. Consequently, he created his
index () based on a study. The study focused on the
skin color change due to UV-irradiation induced
erythema () versus a control () region. Tronnier,
unlike Diffey, accounted for melanin in his
computation. However, he assumed that the relative
calculation should compensate for the melanin
absorption in the skin tissue without the need for a
separate calculation step. In summary, he used the
back reflected light measurements at two wavelengths:
545 nm and 661 nm for both the erythematous (
) and the control regions ( ) to assess the
skin’s relative erythema, as shown in (11).
 
4.4 Ferguson-Hemoglobin Content
Ferguson developed another computation technique
for skin erythema called the index of hemoglobin
content (IHB) [152]. The developed technique
depends on the use of the isosbestic absorption points
of hemoglobin both components (i.e. spectral points of
equal absorption for the oxygenated and deoxygenated
components of hemoglobin). These wavelengths are
the core input data for IHB computation. The term
ODxxx is the optical density of the skin at certain
wavelengths of xxx nanometers.
  
 
 
4.5 Hajizadeh Method
Hajizadeh conducted two studies [153], [68] taking
into consideration the melanin absorption effect on the
measurement of the backscattered light from the
epidermis layer. The two studies paid more attention
to highly pigmented participants versus light skinned
participants. To this end, the melanin index  was
computed as shown in (13) based on the absorption
curve region, similarly to Dawson’s method [154].
However, Hajizadeh had proposed a new
compensation factor due to a synthetic melanin
absorption measurement. This measurement is
equivalent to the in vivo one, and thus he had
developed the term named after him, which is
expressed in µg/cm2 as shown in (14).
 
 
  
Using the newly developed melanin index, , the
“true” hemoglobin content can be computed. This is
done by applying an empirical constant as shown in
  
4.6 Helen Hayes Hospital
Helen Hayes Hospital (HHH) method is an approach
of quantifying erythema mentioned elsewhere [148].
HHH uses an in vitro measurement for a standard
melanin sample as a reference. The reference sample
absorbance is tunable to match the individuals’ skin
melanin content within the spectral range 500-625 nm.
This method has the advantage of consistency for both
high and low pigmented skin. However, this method
has a shortcoming due to its impotence to
physiologically interpret the melanin content changes
between the different locations in the same person, for
example between an erythematous and a control skin
region [148].
5 Discussion
No doubt, VA is still the gold standard [97] and the
primary method for skin erythema assessment. It has
not yet been supplanted by any other methods . As a
proof, a study [155] provided evidence of imprecise
spectrometer results compared to the clinical visual
assessment. Figure 17, belongs to this study, and
shows that the erythema index (d IE) is inconsistent
with a clinical assessment of formaldehyde-induced
erythema [155]. It is of interest to mention that VA is
not perfect since it has critical shortcomings.
Primarily, VA is based on a subjective perspective,
and thus suffers from inter- and intra-observation
individuality, as well as dependency on scene
conditions. Secondly, the therapist eye is capable of
detecting the difference between the skin colors, but it
is unqualified to designate an absolute value for the
detected color difference. Moreover, VA is poor in
prescisely delineating the margins between distinct
erythematous levels. Another point of critique against
VA is color blindness, which may exist undeclared
among clinicians [156].
Figure 17: Comparison between the Dermaspectrometer delta
erythema index (d IE) and the clinical assessment of formaldehyde
patches applied to patients’ skin. The term d IE is equivalent to the
difference between the erythema reading of the Dermaspectrometer
acquired for ROI and a control region. The horizontal axis is the
clinical skin erythema score which ranges from 0 for normal skin
and +++ for severe skin erythema. The vertical axis is the erythema
index score. The term (n) is the number of people who have the same
reaction, the (*) sign is indicating high correlation with positive
patch test (P<0.05, ANOVA test), and the sign () is pointing to
separable categories (P<0.05, ANOVA test) [155].
To overcome the low precision of the therapist eye,
inexpensive and portable color charts are used as a
supplementary tool. They help the clinician to
document the color tone of the skin, a quality that is
hard to remember and to communicate verbally.
Although color charts reduced the limitation of VA,
they still suffer from individualism and durability. In
addition, color charts are vulnerable to metamerism,
which is the inconsistent matching of two colors as a
result of certain conditions including illumination
type, geometrical position, and human-to-human color
detection acuity.
Going far from subjective means, tristimulus
instruments opened the gate for objective assessment
for skin pigmentation. However, these devices are
heavy and pricey as well. Hence, these devices failed
to achieve popularity in clinics. In terms of size and
weight, the narrowband devices, Mexameter and
Dermaspectrometer, garnered some interest from
clinicians. Nevertheless, the narrowband devices are
poor at distinguishing between different skin types
since their operation is dependent on limited
wavelengths. One more limitation of the former
devices is the need for skin contact. This contact does
not only limit their use in sensitive regions including
burns, but also renders them susceptible to stuck with
local skin variations due to the limited region of
The spectral limitation of the tristimulus, and the
narrowband devices was the reason behind bringing
diffuse reflectance measurement into skin assessment.
DRS measurement techniques consequently were
utilized to detect changes in skin reflectance as a result
of cutaneous disease or treatment. Although DRS
measurement techniques are efficient in determining
the optical properties of tissues, they still have a few
limitations. For instance, the typical IS-DRS
configuration has less precision in the spatial domain.
Hence IS-DRS becomes challenging in separating the
particular influence of scattering and absorption in the
absence of a predetermined scattering spectrum for the
sample under test [85]. Like IS-DRS, the spatially-
resolved DRS is not flawless. It is dependent on a few
local point measurements which might yield
indecisive outcomes [157]. Far from CW illumination,
TD-DRS measurement techniques provide the
advantage of acquiring both absorption and scattering
coefficients data via a single measurement.
Nonetheless, it requires a complex set of pricey
instrumentation, which is not available everywhere.
The prior DRS techniques disadvantages paved the
way for imaging techniques to advance toward skin
erythema quantitative assessment.
Digital imaging use in erythema assessment is a tough
process because it needs firm constraints in order to
obtain informative data. To give a brief, the
illumination in digital imaging needs to be securely
maintained within the same level of intensity, emitted
spectra. For illustration purposes, Figure 18 [91]
displays two non-identical pictures of a single patient
taken at two different illumination conditions.
Although there is no change except in the illumination,
the two pictures may be falsely interpreted by the
clinician as two distinct skin conditions.
Figure 18: Example of illumination conditions impact on the
outcome of photographed pictures, which may lead to an error in
clinical diagnosis [91].
To overcome both illumination problems and specular
glares, polarized imaging was solicited. However,
polarized imaging requires an integration of additional
optical components. Such addition of the optical
components leads to: first, an increase in the cost of
the developed optical setup, and second, growth the
system’s sheer complexity with respect to setup and
alignment. Besides, the greater the number of the
integrated optical components in optical imaging, the
more intense the illumination and the more sensitive
the detector are required.
Although fluorescence imaging has a great potential to
display skin structure in the molecular scale, its use is
limited due to the possible harm of UV light to the eyes
and the skin. In addition, fluorescence imaging
requires special tools/ instrumentations, and personal
protective equipment for both the patient and the
imaging operator, which constitute an overhead cost.
For instance, high sensitive detectors are required to
detect the low intensities of both the exciting
illumination and the fluorescence emission since the
band filters or a monochromator are used in excitation
and emission.
The microscope is the shortcut technique to extract the
root information about skin diseases associated with
erythema. Despite this fact, the microscope is not
popular since it is expensive, and difficult to move.
Moreover, the microscope often requires a skin biopsy
for laboratory analysis. Hence, a change of the optical
properties may take place during transport of the
biopsy sample from the clinic to the laboratory.
Coupled with the relatively long time required for lab
analysis, microscopy is limited to the analysis of small
skin regions. Far from common techniques, OCT,
ultrasound, MRI, and dielectric constant
measurements, are less attractive for dermatology
clinics, due to many reasons such as their size, cost,
and inconvenience.
The microscope is the shortcut technique to extract the
root information about diseases associated with
erythematous signs. Nontheless, the costs and limited
portability of microscopes yielded no popularity in
skin erythema assessment. Moreover, the microscope
often requires skin biopsy for laboratory analysis.
Hence, a change of optical properties may take place
during transport of the biopsy sample from the clinic
to the laboratory. Coupled with the relatively long time
necessary for the lab analysis, microscopy is limited to
the analysis of small skin regions.
Far from common techniques, OCT, ultrasound, MRI,
and dielectric constant measurements, are the least
attractive and least common, methods inthe
dermatology clinics, due to many reasons such as the
size, cost, and inconvenience.
Aside from common and less common erythema
assessement techniques, spectral imaging could be an
effective assessment candidate. Spectral imaging is a
hybrid imaging technique.It combines both DRS and
digital imaging, and thus it overcomes the limitations
of each individual technique. For instance, HSI is
tagged as the gold standard in skin erythema mapping
due its distinguished capability of contouring the
distinct skin response intensities. Furthermore, HSI
datacube analysis enables a precise estimation of the
temporal changes of the skin’s componential
chromophores. However, HSI still requires a
reasonable budget, for either case, building a custom-
made spectral camera or purchasing a bench-top one.
In addition to the required budget for HSI, one more
challenge is to develop an intellectual image
processing algorithm to achieve an efficient image
registration and precise endmembers classification.
6 Conclusion
Optical science and technology have resulted in
incredible advances in miniaturizing apparatus and
full-systems to facilitate skin erythema quantification.
Handheld colorimeters, single point spectroscopic
devices, digital and spectral cameras are now mobile,
pocket-sized, and ready to offer objective measures for
various skin diseases associated with erythema. Larger
imaging devices are equipped with high definition
monitors, capable of producing real-time false color
images and displaying the skin’s epidermal and dermal
symptoms. More sophisticated methods, such as
confocal microscopy and OCT, offer almost
histological spatial resolutions for details of the
erythematous skin. This versatile array of objective
tools to study cutaneous diseases associated with skin
erythema enhances clinical dermatology practice, in
terms of both sensitivity and specificity.
However, the capabilities of the expert clinician’s eyes
remains irrreplaceable for skin erythema inspection.
Nevertheless, technology offers a few solutions for
eliminating subjectivity and addressing the previously
mentioned disadvantages of VA. From an economic
perspective, hiring a number of full-time
radiotherapists might be more expensive than
purchasing a technology in the long run. Therefore,
clinicians who are equipped with an advanced
technology are able to provide good service for more
patients in a shorter peroid of time. From our point of
view, HSI, as a developing optical modality, is a
promising future candidate to facilitating the work of
clinicians. Since HSI is an objective, precise, and
contactless approach to skin pigmentation mapping for
a sizable tissue region.
[1] M. Nischik and C. Forster, “Analysis of skin erythema
using true-color images.,” IEEE Trans. Med. Imaging,
vol. 16, no. 6, pp. 7116, 1997.
[2] G. Marchini, A. Nelson, J. Edner, S. Lonne-Rahm, A.
Stavreus-Evers, and K. Hultenby, “Erythema Toxicum
Neonatorum Is an Innate Immune Response to
Commensal Microbes Penetrated into the Skin of the
Newborn Infant,” Pediatr. Res., vol. 58, no. 3, pp. 613
616, 2005.
[3] S. L. Brice et al., “The herpes-specific immune
response of individuals with herpes-associated
erythema multiforme compared with that of
individuals with recurrent herpes labialis,” Arch
Dermatol Res, vol. 285, pp. 193196, 1993.
Quat. J. Exp. Physiol, vol. 6, pp. 339354, 1913.
[5] H. M. M. T. Lewis, R. T. Grant, “Vascular reactions of
the skin to injury, Part X.The intervention of a
chemical stimulus illustrated especially by the flare.
The response to faradism,” Heart, vol. 14, pp. 139
160, 1927.
[6] H. Hönigsmann, “Erythema and pigmentation,”
Photodermatol. Photoimmunol. Photomed., vol. 18, pp.
7581, 2002.
[7] K.-S. Suh et al., “A long-term evaluation of erythema
and pigmentation induced by ultraviolet radiations of
different wavelengths.,” Ski. Res. Technol., vol. 13, no.
4, pp. 3608, 2007.
[8] J. W. Shin, S. W. Yoon, J. B. Jeong, and K. C. Park,
“Different responses of the melanin index to
ultraviolet irradiation in relation to skin color and
body site,” Photodermatol. Photoimmunol. Photomed.,
vol. 30, no. 6, pp. 308315, 2014.
[9] M. Bodekær, P. A. Philipsen, T. Karlsmark, and H. C.
Wulf, “Good agreement between minimal erythema
dose test reactions and objective measurements: An
in vivo study of human skin,” Photodermatol.
Photoimmunol. Photomed., vol. 29, no. 4, pp. 190195,
[10] L. R. Sklar, F. Almutawa, H. W. Lim, and I. Hamzavi,
“Effects of ultraviolet radiation, visible light, and
infrared radiation on erythema and pigmentation: a
review,” Photochem. Photobiol. Sci., vol. 12, no. 1, pp.
5464, 2013.
[11] B. Jung, S. Kim, Y. Bae, H. Kang, Y. Lee, and J. S. Nelson,
“Real-time measurement of skin erythema variation
by negative compression: pilot study.,” J. Biomed. Opt.,
vol. 17, no. 8, pp. 814221, Aug. 2012.
[12] L. Kong et al., “Single sensor that outputs narrowband
multispectral images.,” J. Biomed. Opt., vol. 15, no. 1, p.
10502, 2010.
[13] M. Denstedt, B. S. Pukstad, L. a. Paluchowski, J. E.
Hernandez-Palacios, and L. L. Randeberg,
“Hyperspectral imaging as a diagnostic tool for
chronic skin ulcers,” vol. 8565, p. 85650N, Mar. 2013.
[14] G. Heyer, O. P. Hornstein, and H. O. Handwerker, “Skin
reactions and itch sensation induced by epicutaneous
histamine application in atopic dermatitis and
controls,” J Invest Dermatol, vol. 93, no. 4, pp. 492
496, 1989.
[15] W. Magerl, R. a Westerman, B. Möhner, and H. O.
Handwerker, “Properties of transdermal histamine
iontophoresis: differential effects of season, gender,
and body region.,” J. Invest. Dermatol., vol. 94, no. 3,
pp. 347352, 1990.
[16] D. E. McKee, D. H. Lalonde, A. Thoma, D. L. Glennie,
and J. E. Hayward, “Optimal Time Delay between
Epinephrine Injection and Incision to Minimize
Bleeding,” Plast. Reconstr. Surg., vol. 131, no. 4, pp.
811814, 2013.
[17] D. L. Glennie, J. E. Hayward, D. E. Mckee, and T. J.
Farrell, “Inexpensive diffuse reflectance spectroscopy
system for measuring changes in tissue optical
properties,” J. Biomed. Opt., vol. 19, no. December, pp.
105005-1-1050056, 2014.
[18] E. Van Den Kerckhove et al., “The assessment of
erythema and thickness on burn related scars during
pressure garment therapy as a preventive measure
for hypertrophic scarring,” Burns, vol. 31, no. 6, pp.
696702, 2005.
[19] S. D’haese, M. Van Roy, T. Bate, P. Bijdekerke, and V.
Vinh-Hung, “Management of skin reactions during
radiotherapy in Flanders (Belgium): a study of
nursing practice before and after the introduction of a
skin care protocol.,” Eur. J. Oncol. Nurs., vol. 14, no. 5,
pp. 367372, 2010.
[20] D. Yohan et al., “Quantitative monitoring of radiation
induced skin toxicities in nude mice using optical
biomarkers measured from diffuse optical reflectance
spectroscopy.,” Biomed. Opt. Express, vol. 5, no. 5, pp.
130920, 2014.
[21] L. Diana L. Glennie, Joseph E. Hayward, Orest Z.
Ostapiak, James Wright and T. J. F. Doerwald-Munoz,
“Diffuse reflectance spectroscopy for monitoring
erythema in head & neck intensity modulated
radiation therapy,” J. Radiat. Oncol., 2014.
[22] N. Lee et al., “Skin toxicity due to intensity-modulated
radiotherapy for head-and-neck carcinoma,” Int. J.
Radiat. Oncol. Biol. Phys., vol. 53, no. 3, pp. 630637,
[23] D. Porock and L. Kristjanson, “Skin reactions during
radiotherapy for breast cancer: The use and impact of
topical agents and dressings,” Eur. J. Cancer Care
(Engl)., vol. 8, no. 3, pp. 143153, 1999.
[24] M. McQuestion, “Evidence-Based Skin Care
Management in Radiation Therapy,” Semin. Oncol.
Nurs., vol. 22, no. 3, pp. 163173, 2006.
[25] J. W. Denham et al., “Factors influencing the degree of
erythematous skin reactions in humans,” Radiother.
Oncol., vol. 36, no. 2, pp. 107120, 1995.
[26] T. Olschewski, K. Bajor, B. Lang, E. Lang, and M. H.
Seegenschmiedt, “[Radiotherapy of basal cell
carcinoma of the face and head: Importance of low
dose per fraction on long-term outcome].,” J. Dtsch.
Dermatol. Ges., vol. 4, no. 2, pp. 12430, 2006.
[27] D. Yohan et al., “Quantitative monitoring of radiation
induced skin toxicities in nude mice using optical
biomarkers measured from diffuse optical reflectance
spectroscopy.,” Biomed. Opt. Express, vol. 5, no. 5, pp.
130920, 2014.
[28] D. Basketter, F. Reynolds, M. Rowson, C. Talbot, and E.
Whittle, “Visual assessment of human skin irritation:
a sensitive and reproducible tool.,” Contact
Dermatitis, vol. 37, pp. 218220, 1997.
[29] M. Bruze, M. Isaksson, B. Edman, B. Björkner, S.
Fregert, and H. Möller, “A study on expert reading of
patch test reactions: inter-individual accordance,”
Contact Dermatitis, vol. 32, pp. 331337, 1995.
[30] E. Held, H. Lorentzen, T. Agner, and T. Menné,
“Comparison between visual score and erythema
index (DermaSpectrometer) in evaluation of allergic
patch tests,” Ski. Res. Technol., vol. 4, no. 4, pp. 188
191, 1998.
[31] S. Wan, J. a Parrish, and K. F. Jaenicke, “Quantitative
evaluation of ultraviolet induced erythema.,”
Photochem. Photobiol., vol. 37, no. 6, pp. 643648,
[32] B. L. Diffey and P. M. Farr, “Quantitative aspects of
ultraviolet erythema,” Clin. Phys. Physiol. Meas., vol.
12, no. 4, p. 311, 1991.
[33] A. S. C. J. Taylor S.C., “The Taylor Hyperpigmentation
Scale: A New Visual Assessment Tool for the
Evaluation of Skin Color and Pigmentation,” Ther.
Clin., vol. 76, no. 4, pp. 270274, 2005.
[34] M. H. Ahmad Fadzil, D. Ihtatho, A. Mohd Affandi, and
S. H. Hussein, “Objective assessment of psoriasis
erythema for PASI scoring.,” J. Med. Eng. Technol., vol.
33, no. 7, pp. 51624, 2009.
[35] S. Banu, G. Toacse, and G. Danciu, “Objective erythema
assessment of Psoriasis lesions for Psoriasis Area and
Severity Index (PASI) evaluation,” EPE 2014 - Proc.
2014 Int. Conf. Expo. Electr. Power Eng., no. Epe, pp.
5256, 2014.
[36] A. Raina, R. Hennessy, M. Rains, J. Allred, D. Diven, and
M. K. Markey, “Objective measurement of erythema in
psoriasis using digital color photography with color
calibration,” in Annual International Conference of the
IEEE Engineering in Medicine and Biology Society.,
2014, vol. 2014, pp. 33333336.
[37] N. Kollias and G. N. Stamatas, “Optical non-invasive
approaches to diagnosis of skin diseases,” J. Investig.
Dermatology Symp. Proc., vol. 7, pp. 6475, 2002.
[38] R. M. M. h. A. Lahti, H. Kopola, A. Harila, “Assessment
of skin erythema by eye, laser doppler flowmeter,
spectroradiometer, two channel erythema meter, and
Minolta chromameter,” Dermatological Res., pp. 278
282, 1993.
[39] and W. I. l. H. E. L. M. V. Hausser, karl Wilhelm,
Sunburn and suntanning. London: Pergamon press,
[40] S. Taylor, W. Westerhof, S. Im, and J. Lim,
Noninvasive techniques for the evaluation of skin
color,” J. Am. Acad. Dermatol., vol. 54, no. 5, pp. S282
S290, 2006.
[41] I. M. Gibson, “Measurement of skin colour in vivo,” J.
Soc. Cosmet. Chem, vol. 22, pp. 725740, 1971.
[42] N. Konishi et al., “New approach to the evaluation of
skin color of pigmentary lesions using Skin Tone
Color Scale.,” J. Dermatol., vol. 34, no. 7, pp. 441446,
[43] R. G. Kuehni, Color space and its divisions, vol. 26, no.
3. Canada: John Wiley & Sons, Inc., Hoboken, New
Jersey, 2003.
[44] I. L. Weatherall and B. D. Coombs, “Skin color
measurements in terms of CIELAB color space
values.,” J. Invest. Dermatol., vol. 99, no. 4, pp. 468
473, 1992.
[45] M. R., Color physics for Industry. Society of Dyers and
Colorists, Bradford, 1987.
[46] F. R. A. S. T. SMITH, M.A., F.INsT.P., J. GUILD, A.R.C.S.,
AND THEIR USE,” Trans. Opt. Soc., vol. XXXIII, no. 3,
[47] Daniel Malacara, “Color Vision and Colorimetry
theory and Applications,” 2nd ed., Society of Photo-
Optical Instrumentation Engineers (SPIE), 2011.
[48] R. P. C. O. Crick et al., A TEXTBOOK of CLINICAL
OPHTHALMOLOGY. World Scientific Publishing Co.
Pte. Ltd., 2003.
[49] T. Root, Ophthobook. CreateSpace Independent
Publishing Platform, 2009.
[50] J. Serup and T. Agner, “Colorimetric quantification of
erythema-a comparison of two colorimeters (Lange
Micro Color and Minolta Chroma Meter CR-200) with
a clinical scoring scheme and laser-Doppler
flowmetry,” Clin. Exp. Dermatol., vol. 15, no. 4, pp.
267272, Jul. 1990.
[51] C. Queille-Roussel, M. Poncet, and H. Schaefer,
“Quantification of skin-colour changes induced by
topical corticosteroid preparations using the Minolta
Chroma Meter,” Br J Dermatol, vol. 124, no. 3, pp.
264270, 1991.
[52] A. Fullerton, T. Fischer, A. Lahti, K. P. Wilhelm, H.
Takiwaki, and J. Serup, “Guidelines for measurement
of skin colour and erythema. A report from the
Standardization Group of the European Society of
Contact Dermatitis.,” Contact dermatitis, vol. 35, no. 1.
pp. 110, 1996.
[53] E. Van den Kerckhove, F. Staes, M. Flour, K.
Stappaerts, and W. Boeckx, “Reproducibility of
repeated measurements on healthy skin with Minolta
Chromameter CR-300.,” Skin Res. Technol., vol. 7, no.
1, pp. 569, 2001.
[54] S. Park, C. Huh, Y. Choe, and J. Youn, “Time course of
ultraviolet induced skin reactions evaluated by two
different reflectance spectrophotometers :
DermaSpectrophotometer A and Minolta
spectrometer CM-2002,” Photodermatol.
Photoimmunol. Photomed., vol. 18, pp. 2328, 2002.
[55] C. K. Kraemer, D. B. Menegon, and T. F. Cestari,
“Determination of the minimal phototoxic dose and
colorimetry in psoralen plus ultraviolet A radiation
therapy,” Photodermatol. Photoimmunol. Photomed.,
vol. 21, no. 5, pp. 242248, 2005.
[56] M. D. Shriver and E. J. Parra, “Comparison of narrow-
band reflectance spectroscopy and tristimulus
colorimetry for measurements of skin and hair color
in persons of different biological ancestry,” Am. J.
Phys. Anthropol., vol. 112, no. 1, pp. 1727, 2000.
[57] L. J. Draaijers, F. R. . Tempelman, Y. A. . Botman, R. W.
Kreis, E. Middelkoop, and P. P. . van Zuijlen, “Colour
evaluation in scars: tristimulus colorimeter, narrow-
band simple reflectance meter or subjective
evaluation?,” Burns, vol. 30, no. 2, pp. 103107, 2004.
[58] M. Baquie and B. Kasraee, “Discrimination between
cutaneous pigmentation and erythema: Comparison
of the skin colorimeters Dermacatch and Mexameter,”
Ski. Res. Technol., vol. 20, no. 2, pp. 218227, 2014.
[59] A. R. Matias, M. Ferreira, P. Costa, and P. Neto, “Skin
colour, skin redness and melanin biometric
measurements: Comparison study between Antera-
3D, Mexameter and Colorimeter,” Ski. Res. Technol.,
pp. 346362, 2015.
[60] K. Yoshimura, K. Harii, Y. Masuda, M. Takahashi, T.
Aoyama, and T. Iga, “Usefulness of a narrow-band
reflectance spectrophotometer in evaluating effects
of depigmenting treatment,” Aesthetic Plast. Surg., vol.
25, no. 2, pp. 129133, 2001.
[61] B. L. Diffey, R. J. Oliver, and P. M. Farr, “A portable
instrument for quantifying erythema induced by
ultraviolet radiation,” Br. J. Dermatol., vol. 111, no. 6,
pp. 663672, Dec. 1984.
[62] P. M. Farr and B. L. Diffey, “Quantitative studies on
cutaneous erythema induced by ultraviolet
radiation.,” Br. J. Dermatol., vol. 111, no. 6, pp. 673
682, 1984.
[63] J. R. Trevithick et al., “Topical tocopherol acetate
reduces post-UVB, sunburn-associated erythema,
edema, and skin sensitivity in hairless mice,” Arch.
Biochem. Biophys., vol. 296, no. 2, pp. 575582, 1992.
[64] s poon t c, j m kuchel, A. Badruddin, and G. M.
Halliday, “Objective measurement of minimal
erythemal and melanogenic doses using natural and
solar simulated light,” Photochem. Photobiol., vol. 78,
no. 4, pp. 331336, 2003.
[65] E. S. Park, J. I. Na, S. O. Kim, C. H. Huh, S. W. Youn, and
K. C. Park, “Application of a pigment measuring
device - Mexameter - for the differential diagnosis of
vitiligo and nevus depigmentosus,” Ski. Res. Technol.,
vol. 12, no. 4, pp. 298302, 2006.
[66] K. S. Cheng, M. W. Huang, and P. Y. Lo, “Objective
assessment of sunburn and minimal erythema doses:
Comparison of noninvasive in vivo measuring
techniques after UVB irradiation,” EURASIP J. Adv.
Signal Process., vol. 2010, pp. 17, 2010.
[67] E. J. Yoshida, H. Chen, M. A. Torres, W. J. Curran, and T.
Liu, “Spectrophotometer and ultrasound evaluation
of late toxicity following breast-cancer radiotherapy.,”
Med. Phys., vol. 38, no. 10, pp. 574755, 2011.
[68] M. Hajizadeh-Saffar, J. W. Feather, and J. B. Dawson,
“An investigation of factors affecting the accuracy of
in vivo measurements of skin pigments by reflectance
spectrophotometry,” Phys. Med. Biol. Phys. Med. Biol,
vol. 35, no. 9, pp. 13011315, 1990.
[69] M. S. Patterson, B. Chance, and B. C. Wilson, “Time
resolved reflectance and transmittance for the non-
invasive measurement of tissue optical properties,”
Appl. Opt., vol. 28, no. 12, pp. 23312336, 1989.
[70] D. T. Delpy, M. Cope, P. van der Zee, S. Arridge, S.
Wray, and J. Wyatt, “Estimation of optical pathlength
through tissue from direct time of flight
measurement.,” Phys. Med. Biol., vol. 33, no. 12, pp.
14331442, 1988.
[71] M. S. Patterson, J. D. Moulton, B. C. Wilson, K. W.
Berndt, and J. R. Lakowicz, “Frequency-domain
reflectance for the determination of the scattering
and absorption properties of tissue,Appl. Opt., vol.
30, no. 31, pp. 44744476, 1991.
[72] M. S. Patterson and B. W. Pogue, “Mathematical model
for time-resolved and frequency-domain fluorescence
spectroscopy in biological tissues,” Appl. Opt., vol. 33,
no. 10, pp. 19631974, 1994.
[73] N. Kollias, I. Seo, and P. R. Bargo, “Interpreting diffuse
reflectance for in vivo skin reactions in terms of
chromophores,” J. Biophotonics, vol. 3, no. 12, pp.
1524, 2010.
[74] L. Lim, B. Nichols, N. Rajaram, and J. W. Tunnell,
“Probe pressure effects on human skin diffuse
reflectance and fluorescence spectroscopy
measurements,” J. Biomed. Opt., vol. 16, no. 1, p.
11012, 2011.
[75] J. R. Mallia, S. S. Thomas, and P. Sebastian, “Oral
cancer detection using diffuse reflectance spectral
ratio R540 / R575 of oxygenated hemoglobin bands,”
vol. 11, no. February 2006, pp. 16, 2017.
[76] A. H. Taylor, “The measurement of diffuse reflection
factors and a new absolute reflectometer,” J. Opt. Soc.
Am., vol. 4, no. 12, p. 9, 1920.
[77] J. a. Jacquez and H. F. Kuppenheim, “Theory of the
Integrating Sphere,” J. Opt. Soc. Am., vol. 45, no. 6, p.
460, 1955.
[78] J. W. Pickering, C. J. M. Moes, H. J. C. M. Sterenborg, S.
A. Prahl, and M. J. C. van Gemert, “Two integrating
spheres with an intervening scattering sample,” J. Opt.
Soc. Am. A, vol. 9, no. 4, pp. 621631, 1992.
[79] J. W. Pickering, S. a Prahl, N. van Wieringen, J. F. Beek,
H. J. Sterenborg, and M. J. van Gemert, “Double-
integrating-sphere system for measuring the optical
properties of tissue.,” Appl. Opt., vol. 32, no. 4, pp.
399410, 1993.
[80] T. J. Farrell, “A diffusion theory model of spatially
resolved, steady-state diffuse reflectance for the
noninvasive determination of tissue optical
properties in vivo,” Med. Phys., vol. 19, no. 4, p. 879,
[81] a Kienle, L. Lilge, M. S. Patterson, R. Hibst, R. Steiner,
and B. C. Wilson, “Spatially resolved absolute diffuse
reflectance measurements for noninvasive
determination of the optical scattering and
absorption coefficients of biological tissue.,” Appl.
Opt., vol. 35, no. 13, pp. 23042314, 1996.
[82] R. M. Doornbos, R. Lang, M. C. Aalders, F. W. Cross,
and H. J. Sterenborg, “The determination of in vivo
human tissue optical properties and absolute
chromophore concentrations using spatially resolved
steady-state diffuse reflectance spectroscopy.,” Phys.
Med. Biol., vol. 44, no. 4, pp. 967981, 1999.
[83] A. Pifferi et al., “Spectroscopic time-resolved diffuse
reflectance and transmittance measurements of the
female breast at different interfiber distances.,” J.
Biomed. Opt., vol. 9, no. 6, pp. 114351, 2004.
[84] D. F. Moscu, “Quantifying chromophore concentration
in tissue simulating phantoms using an optical
detection system based on an integrating sphere,” p.
153, 2010.
[85] D. F. Moscu, “Quantifying chromophore concentration
in tissue simulating phantoms using an optical
detection system based on an integrating sphere,”
[86] K. Wårdell, H. K. Naver, G. E. Nilsson, and B. G. Wallin,
“The cutaneous vascular axon reflex in humans
characterized by laser Doppler perfusion imaging.,” J.
Physiol., vol. 460, pp. 185199, 1993.
[87] G. Heyer, O. P. Hornstein, and H. O. Handwerker, “Skin
reactions and itch sensation induced by epicutaneous
histamine application in atopic dermatitis and
controls,” J Invest Dermatol, vol. 93, no. 4, pp. 492
496, 1989.
[88] G. V Oliveira, D. Chinkes, C. Mitchell, G. Oliveras, H. K.
Hawkins, and D. N. Herndon, “Objective assessment of
burn scar vascularity, erythema, pliability, thickness,
and planimetry.,” Dermatol. Surg., vol. 31, pp. 4858,
[89] S. B. Phillips, N. Kollias, R. Gillies, J. A. Muccini, and L.
A. Drake, “Polarized light photography enhances
visualization of inflammatory lesions of acne
vulgaris,” J. Am. Acad. Dermatol., vol. 37, no. 6, pp.
948952, 1997.
[90] A. C. Halpern, A. A. Marghoob, T. W. Bialoglow, W.
Witmer, and W. Slue, “Standardized positioning of
patients (poses) for whole body cutaneous
photography,” J. Am. Acad. Dermatol., vol. 49, no. 4, pp.
593598, 2003.
[91] W. K. Witmer and P. J. Lebovitz, “Clinical Photography
in the Dermatology Practice,” Semin. Cutan. Med.
Surg., vol. 31, no. 3, pp. 191199, 2012.
[92] A. Raina, R. Hennessy, M. Rains, J. Allred, D. Diven, and
M. K. Markey, “Objective measurement of erythema in
psoriasis using digital color photography with color
calibration,” Ski. Res. Technol., no. 7, pp. 16, 2015.
[93] M. Setaro and A. Sparavigna, “Quantification of
erythema using digital camera and computer-based
colour image analysis: a multicentre study.,” Skin Res.
Technol., vol. 8, no. 2, pp. 848, 2002.
[94] D. Hossam, A. Sadek, A. Wagih, and M. Malik, “Clinical
photography of skin diseases,” Egypt. Dermatology
Online J., vol. 11, no. 1, pp. 110, 2015.
[95] “CANFIELD.” [Online]. Available:
[96] S. G. Coelho, S. A. Miller, B. Z. Zmudzka, and J. Z. Beer,
“Quantification of UV-Induced Erythema and
Pigmentation Using Computer-Assisted Digital Image
Evaluation,” Photochem. Photobiol., vol. 82, no. 3, pp.
651655, 2006.
[97] S. Taylor, W. Westerhof, S. Im, and J. Lim,
“Noninvasive techniques for the evaluation of skin
color,” J. Am. Acad. Dermatol., vol. 54, no. 5, pp. S282
S290, 2006.
[98] A. K. CC Zouboulis, AD Katsambas, Pathogenesis and
Treatment of Acne and Rosacea. Springer-Verlag
Berlin Heidelberg, 2014.
[99] S. Sridhar and A. Da Silva, “Enhanced contrast and
depth resolution in polarization imaging using
elliptically polarized light,” J. Biomed. Opt., vol. 21, no.
7, p. 71107, 2016.
[100] Z. Huang et al., “Cutaneous melanin exhibiting
fluorescence emission under near-infrared light
excitation.,” J. Biomed. Opt., vol. 11, no. 3, p. 34010,
[101] J. O’doherty, J. Henricson, C. Anderson, M. J. Leahy, G.
E. Nilsson, and F. Sjoberg, “Sub-epidermal imaging
using polarized light spectroscopy for assessment of
skin microcirculation,” Ski. Res. Technol., vol. 13, no. 4,
pp. 472484, 2007.
[102] J. M. Gallas and M. Eisner, “Fluorescence of Melanin-
Dependence Upon Excitation Wavelength and
Concentration,” Photochem. Photobiol., vol. 45, no. 5,
pp. 595600, 1987.
[103] N. Kollias et al., “Fluorescence photography in the
evaluation of hyperpigmentation in photodamaged
skin,” J. Am. Acad. Dermatol., vol. 36, no. 2, pp. 226
300, 1997.
[104] R. J. T. Leslie C. Lucchina, Scott B. Phillips,
“Fluorescence photography in the evaluation of acne,”
Am. Acad. Dermatology, 1996.
[105] A. Tosti et al., “Follicular Red Dots,” Arch Dermatol,
vol. 145, no. 12, pp. 14061409, 2009.
[106] A. N. Yaroslavsky, J. Barbosa, V. Neel, C. DiMarzio, and
R. R. Anderson, “Combining multispectral polarized
light imaging and confocal microscopy for localization
of nonmelanoma skin cancer.,” J. Biomed. Opt., vol. 10,
no. 1, p. 14011, 2005.
[107] M. Rajadhyaksha, S. González, J. M. Zavislan, R. R.
Anderson, and R. H. Webb, “In vivo confocal scanning
laser microscopy of human skin II: advances in
instrumentation and comparison with histology.,” J.
Invest. Dermatol., vol. 113, no. 3, pp. 293303, 1999.
[108] S. Astner, E. Gonzalez, A. Cheung, F. Rius-Diaz, and S.
González, “Pilot study on the sensitivity and
specificity of in vivo reflectance confocal microscopy
in the diagnosis of allergic contact dermatitis,” J. Am.
Acad. Dermatol., vol. 53, no. 6, pp. 986992, 2005.
[109] S. Astner, N. Burnett, F. Rius-Díaz, A. G. Doukas, S.
González, and E. Gonzalez, “Irritant contact dermatitis
induced by a common household irritant: A
noninvasive evaluation of ethnic variability in skin
response,” J. Am. Acad. Dermatol., vol. 54, no. 3, pp.
458465, 2006.
[110] J. a Suárez-Pérez, “Pathogenesis and diagnosis of
contact dermatitis: Applications of reflectance
confocal microscopy,” World J. Dermatology, vol. 3, no.
3, p. 45, 2014.
[111] R. Proc et al., “Imaging Spectroscopy Using Tunable
Filters: A Review,” Apr. 2000.
[112] D. a Glenar, J. J. Hillman, B. Saif, and J. Bergstralh,
“Acousto-optic imaging spectropolarimetry for
remote sensing.,” Appl. Opt., vol. 33, no. 31, pp. 7412
24, Nov. 1994.
[113] L. Kong et al., “Handheld erythema and bruise
detector,” Proc. SPIE, vol. 6915, p. 69153K69153K7,
[114] D. Jakovels, J. Spigulis, and I. Saknite, “Multi-spectral
mapping of in vivo skin hemoglobin and melanin,”
SPIE Proc., vol. 7715, p. 77152Z--77152Z, 2010.
[115] J. M. Beach, M. a. Lanoue, K. Brabham, and B.
Khoobehi, “Portable hyperspectral imager for
assessment of skin disorders: preliminary
measurements,” in Photonic Therapeutics and
Diagnostics, Proc. of SPIE, 2005, vol. 5686, pp. 111
[116] Z. Nie, R. An, J. E. Hayward, T. J. Farrell, and Q. Fang,
“Hyperspectral fluorescence lifetime imaging for
optical biopsy.,” J. Biomed. Opt., vol. 18, no. 9, p.
96001, Sep. 2013.
[117] C. Zhang and H. Wang, “The narrow band AOTF based
hyperspectral microscopic imaging on the rat skin
stratum configuration,” J.Europ.Opt.Soc., vol. 14034,
[118] a. D. Meigs, I. Otten, L.J., and T. Y. Cherezova,
“Ultraspectral imaging: a new contribution to global
virtual presence,” in 1998 IEEE Aerospace Conference
Proceedings (Cat. No.98TH8339), 1998, vol. 2, no.
[119] A. F. H. Goetz, “Three decades of hyperspectral
remote sensing of the Earth: A personal view,”
Remote Sens. Environ., vol. 113, pp. S5S16, 2009.
[120] G. Lu and B. Fei, “Medical hyperspectral imaging: a
review.,” J. Biomed. Opt., vol. 19, no. 1, p. 10901, Jan.
[121] N. Kollias, R. Gillies, J. A. Muccini, R. K. Uyeyama, S. B.
Phillips, and L. A. Drake, “A Single Parameter,
Oxygenated Hemoglobin, Can Be Used to Quantify
Experimental Irritant-Induced Inflammation,” J.
Invest. Dermatol., vol. 104, no. 3, pp. 421424, 1995.
[122] G. N. Stamatas, C. J. Balas, and N. Kollias,
“Hyperspectral Image Acquisition and Analysis of
Skin,” in Spectral Imaging: Instrumentation,
Applications, and Analysis II, Proceedings of SPIE,
2003, vol. 4959, pp. 7782.
[123] G. N. Stamatas, M. Southall, and N. Kollias, “In vivo
monitoring of cutaneous edema using spectral
imaging in the visible and near infrared.,” J. Invest.
Dermatol., vol. 126, no. 8, pp. 17531760, 2006.
[124] G. N. Stamatas and N. Kollias, “In vivo documentation
of cutaneous inflammation using spectral imaging.,” J.
Biomed. Opt., vol. 12, no. 5, p. 51603, 2007.
[125] M. S. Chin et al., “Hyperspectral imaging for early
detection of oxygenation and perfusion changes in
irradiated skin,” J. Biomed. Opt., vol. 17, no. 2, p.
26010, 2012.
[126] G. V. G. Baranoski, S. Member, T. F. Chen, and S.
Member, “On the Identification and Interpretation of
Human Skin Spectral Responses Under Adverse
Environmental Conditions,” pp. 845–848, 2015.
[127] G. N. Stamatas and N. Kollias, “In vivo documentation
of cutaneous inflammation using spectral imaging.,” J.
Biomed. Opt., vol. 12, no. 5, p. 51603, 2007.
[128] S. Zhang, P. Liu, J. Huang, and R. Xu, “Multiview
hyperspectral topography of tissue structural and
functional characteristics,” J. Biomed. Opt., vol. 21, p.
85530P85530P7, 2016.
[129] S. Zhang, P. Liu, J. Huang, and R. Xu, “Multiview
hyperspectral topography of tissue structural and
functional characteristics,” J. Biomed. Opt., vol. 21, p.
85530P85530P7, 2016.
[130] J. Nyström, P. Geladi, B. Lindholm-Sethson, J. Rattfelt,
A. C. Svensk, and L. Franzen, “Objective
measurements of radiotherapy-induced erythema,”
Ski. Res. Technol., vol. 10, no. 4, pp. 242250, 2004.
[131] J. O’doherty, J. Henricson, C. Anderson, M. J. Leahy, G.
E. Nilsson, and F. Sj??berg, “Sub-epidermal imaging
using polarized light spectroscopy for assessment of
skin microcirculation,” Ski. Res. Technol., vol. 13, no. 4,
pp. 472484, 2007.
[132] A. K. CC Zouboulis, AD Katsambas, Pathogenesis and
Treatment of Acne and Rosacea. Springer-Verlag
Berlin Heidelberg, 2014.
[133] M. Boone, G. B. E. Jemec, and V. Del Marmol, “High-
definition optical coherence tomography enables
visualization of individual cells in healthy skin:
Comparison to reflectance confocal microscopy,” Exp.
Dermatol., vol. 21, no. 10, pp. 740744, 2012.
[134] T. Gambichler, G. Moussa, M. Sand, D. Sand, P.
Altmeyer, and K. Hoffmann, “Applications of optical
coherence tomography in dermatology,” J. Dermatol.
Sci., vol. 40, no. 2, pp. 8594, 2005.
[135] J. W. Elke Sattler, Raphaela Kastle, “Optical coherence
tomography in dermatology,” J. Biomed. Opt., vol. 18,
no. 6, 2013.
[136] D. L. M. Harold Alexander, “Determining Skin
Thickness with Pulsed Ultra Sound,” J. Invest.
Dermatol., vol. 72, no. 1, pp. 1719, 1979.
[137] A. Warszawski, E. M. Röttinger, R. Vogel, and N.
Warszawski, “20 MHz ultrasonic imaging for
quantitative assessment and documentation of early
and late postradiation skin reactions in breast cancer
patients,” Radiother. Oncol., vol. 47, no. 3, pp. 241
247, 1998.
[138] E. Szymańska et al., “Skin imaging with high
frequency ultrasound - preliminary results.,” Eur. J.
Ultrasound, vol. 12, pp. 916, 2000.
[139] X. Wortsman and J. Wortsman, “Clinical usefulness of
variable-frequency ultrasound in localized lesions of
the skin,” J. Am. Acad. Dermatol., vol. 62, no. 2, pp.
247256, 2010.
[140] S. Wong, A. Kaur, M. Back, K. M. Lee, S. Baggarley, and
J. J. Lu, “An ultrasonographic evaluation of skin
thickness in breast cancer patients after
postmastectomy radiation therapy.,” Radiat. Oncol.,
vol. 6, no. 1, p. 9, 2011.
[141] X. Wortsman, J. Wortsman, C. Orlandi, G. Cardenas, I.
Sazunic, and G. B. E. Jemec, “Ultrasound detection and
identification of cosmetic fillers in the skin,” J. Eur.
Acad. Dermatology Venereol., vol. 26, no. 3, pp. 292
301, 2012.
[142] R. Kleinerman, T. B. Whang, R. L. Bard, and E. S.
Marmur, “Ultrasound in dermatology: Principles and
applications,” J. Am. Acad. Dermatol., vol. 67, no. 3, pp.
478487, 2012.
[143] X. Wortsman and J. Wortsman, “Clinical usefulness of
variable-frequency ultrasound in localized lesions of
the skin,” J. Am. Acad. Dermatol., vol. 62, no. 2, pp.
247256, 2010.
[144] a B. Kimball et al., “Magnetic resonance imaging
detection of occult skin and subcutaneous
abnormalities in juvenile dermatomyositis.
Implications for diagnosis and therapy.,” Arthritis
Rheum., vol. 43, no. 8, pp. 186673, 2000.
[145] B. Querleux, “Magnetic resonance imaging and
spectroscopy of skin and subcutis,” J Cosmet
Dermatol, vol. 3, no. 3, pp. 156161, 2004.
[146] E. Burdette, F. Cain, and J. Seals, “In vivo probe
measurement technique for determining dielectric
properties at VHF through microwave frequencies,”
IEEE Trans. Microw. Theory Tech., vol. MTT-28, no. 4,
pp. 414427, 1980.
[147] J. Nuutinen et al., “A dielectric method for measuring
early and late reactions in irradiated human skin,”
Radiother. Oncol., vol. 47, no. 3, pp. 249254, 1998.
[148] B. Riordan, S. Sprigle, and M. Linden, “Testing the
validity of erythema detection algorithms.,” J. Rehabil.
Res. Dev., vol. 38, no. 1, pp. 1322, 2001.
[149] I. Diebele, A. Bekina, A. Derjabo, J. Kapostinsh, I.
Kuzmina, and J. Spigulis, “Analysis of skin basalioma
and melanoma by multispectral imaging,”
Biophotonics, vol. 8427, no. May, pp. 842732-842732
5, 2012.
[150] H. Tronnier, “Evaluation and measurement of
ultraviolet erythema,” Biol. Eff. Ultrav. Radiat., pp.
255266, 1969.
[151] B. L. Diffey and P. M. Farr, “Quantitative aspects of
ultraviolet erythema,” Clin. Phys. Physiol. Meas., vol.
12, no. 4, p. 311, 1991.
[152] M. Ferguson-Pell and S. Hagisawa, “An empirical
technique to compensate for melanin when
monitoring skin microcirculation using reflectance
spectrophotometry,” Med. Eng. Phys., vol. 17, no. 2, pp.
104110, 1995.
[153] J. W. Feather, M. Hajizadeh-Saffar, G. Leslie, and J. B.
Dawson, “A portable scanning reflectance
spectrophotometer using visible wavelengths for the
rapid measurement of skin pigments.,” Phys. Med.
Biol., vol. 34, no. 7, pp. 807820, 1989.
[154] J. B. Dawson et al., “A theoretical and experimental
study of light absorption and scattering by in vivo
skin.,” Phys. Med. Biol., vol. 25, no. 4, pp. 695709,
[155] E. Held, H. Lorentzen, T. Agner, and T. Menné,
“Comparison between visual score and erythema
index (DermaSpectrometer) in evaluation of allergic
patch tests,” Ski. Res. Technol., vol. 4, no. 4, pp. 188
191, 1998.
[156] D. M. Cockburn, “Confessions of a colour blind
optometrist,” Clin. Exp. Optom., vol. 87, no. 45, pp.
350352, 2004.
[157] P. R. Bargo et al., “In vivo determination of optical
properties of normal and tumor tissue with white
light reflectance and an empirical light transport
model during endoscopy.,” J. Biomed. Opt., vol. 10, no.
3, p. 34018, 2005.
... To obtain standardized images, a large amount of expensive equipment and expertise is required. In addition, this method requires no contact with the skin and large areas can be evaluated [13]. Hyperspectral imaging combines the advantages of both methods by illuminating the skin and measuring the reflected light at a distance. ...
Full-text available
Although significant advancements in computer-aided diagnostics using artificial intelligence (AI) have been made, to date, no viable method for radiation-induced skin reaction (RISR) analysis and classification is available. The objective of this single-center study was to develop machine learning and deep learning approaches using deep convolutional neural networks (CNNs) for automatic classification of RISRs according to the Common Terminology Criteria for Adverse Events (CTCAE) grading system. ScarletredⓇ Vision, a novel and state-of-the-art digital skin imaging method capable of remote monitoring and objective assessment of acute RISRs was used to convert 2D digital skin images using the CIELAB color space and conduct SEV* measurements. A set of different machine learning and deep convolutional neural network-based algorithms has been explored for the automatic classification of RISRs. A total of 2263 distinct images from 209 patients were analyzed for training and testing the machine learning and CNN algorithms. For a 2-class problem of healthy skin (grade 0) versus erythema (grade ≥ 1), all machine learning models produced an accuracy of above 70%, and the sensitivity and specificity of erythema recognition were 67–72% and 72–83%, respectively. The CNN produced a test accuracy of 74%, sensitivity of 66%, and specificity of 83% for predicting healthy and erythema cases. For the severity grade prediction of a 3-class problem (grade 0 versus 1 versus 2), the overall test accuracy was 60–67%, and the sensitivities were 56–82%, 35–59%, and 65–72%, respectively. For estimating the severity grade of each class, the CNN obtained an accuracy of 73%, 66%, and 82%, respectively. Ensemble learning combines several individual predictions to obtain a better generalization performance. Furthermore, we exploited ensemble learning by deploying a CNN model as a meta-learner. The ensemble CNN based on bagging and majority voting shows an accuracy, sensitivity and specificity of 87%, 90%, and 82% for a 2-class problem, respectively. For a 3-class problem, the ensemble CNN shows an overall accuracy of 66%, while for each grade (0, 1, and 2) accuracies were 76%, 69%, and 87%, sensitivities were 70%, 57%, and 71%, and specificities were 78%, 75%, and 95%, respectively. This study is the first to focus on erythema in radiation-dermatitis and produces benchmark results using machine learning models. The outcome of this study validates that the proposed system can act as a pre-screening and decision support tool for oncologists or patients to provide fast, reliable, and efficient assessment of erythema grading.
... To obtain standardized images, a large amount of expensive equipment and expertise is required. In addition, this method requires no contact with the skin and large areas can be evaluated [13]. Hyperspectral imaging combines the advantages of both methods by illuminating the skin and measuring the reflected light at a distance. ...
Full-text available
Although significant advancements in computer-aided diagnostics using artificial intelligence (AI) have been made, to date, no viable method for radiation-induced skin reaction (RISR) analysis and classification is available. The objective of this single-center study was to develop machine learning and deep learning approaches using deep convolutional neural networks (CNNs) for automatic classification of RISRs according to the Common Terminology Criteria for Adverse Events (CTCAE) grading system. ScarletredⓇ Vision, a novel and state-of-the-art digital skin imaging method capable of remote monitoring and objective assessment of acute RISRs was used to convert 2D digital skin images using the CIELAB color space and conduct SEV* measurements. A set of different machine learning and deep convolutional neural network-based algorithms has been explored for the automatic classification of RISRs. A total of 2263 distinct images from 209 patients were analyzed for training and testing the machine learning and CNN algorithms. For a 2-class problem of healthy skin (grade 0) versus erythema (grade ≥ 1), all machine learning models produced an accuracy of above 70%, and the sensitivity and specificity of erythema recognition were 67-72% and 72-83%, respectively. The CNN produced a test accuracy of 74%, sensitivity of 66%, and specificity of 83% for predicting healthy and erythema cases. For the severity grade prediction of a 3-class problem (grade 0 versus 1 versus 2), the test accuracy was 60-67%, and the sensitivity and specificity were 56-82%, 35-59%, and 65-72%, respectively. For estimating the severity grade of each class, the CNN obtained an accuracy of 73%, 66%, and 82%, respectively. Ensemble learning combines several individual predictions to obtain a better generalization performance. Furthermore, we exploited ensemble learning by deploying a CNN model as a meta-learner. The ensemble CNN based on bagging and majority voting shows an accuracy, sensitivity and specificity of 87%, 90%, and 82% for a 2-class problem, respectively. For a 3-class problem, the ensemble CNN shows an overall accuracy of 66%, while for each grade (0, 1, and 2) accuracies were 0.76%, 0.69%, and 0.87%, sensitivities were 0.70%, 0.57%, and 0.71%, and specificities were 0.78%, 0.75%, and 0.95%, respectively. This study is the first to focus on erythema in radiation-dermatitis and produces benchmark results using machine learning models. The outcome of this study validates that the proposed system can act as a pre-screening and decision support tool for oncologists or patients to provide fast, reliable, and efficient assessment of erythema grading.
Full-text available
Psoriasis severity assessment is usually performed based on the computation of the Psoriasis Area and Severity Index (PASI). Physicians subjectively classify the erythema parameter into several grades of severity. To support the decision and the evaluation of the psoriasis lesions' evolution in time, this study proposes an approach for the objective assessment of erythema degree. There were processed seventeen images depicting psoriasis lesions from mild to severe and the erythema parameter was classified into three categories using machine learning algorithms. The classification was based on color and texture features extracted from the digital images. Three classifiers were trained and tested with these features: Naïve Bayes, Neural Networks and Support Vector Machine. A comparative analysis on the classification accuracy and computation time was made to the classification algorithms. The best classification accuracy (92%) was obtained when using a two-layer feed-forward Neural Network. The proposed method can be used to objectively assess the psoriasis lesion's erythema. The major interest of this approach is to be cheap, fast, robust and easy to use in a dermatological context, with very few constraints on the acquisition protocol.
Full-text available
Contact dermatitis (CD) is the most common professional skin disease, with frequencies ranging from 24 to 170 every 100000 individuals. Approximately 20% of the United States population suffers from CD. CD can be classified according to its origin and severity. ICD stands for irritant CD, whereas ACD means allergic CD. Their clinical presentation includes acute, sub-acute and chronic eczema. Despite their different origin, ICD and ACD often present similar clinical and histologic findings. The current gold standard for diagnosis is patchtesting. However, patch-testing is being questioned in terms of validity and reproducibility, as it relies heavily on the skill of the observer. Real-time reflectance confo cal microscopy is a non-invasive imaging technique that bears strong promise for the study of CD, and it enables the evaluation of cellular and subcellular changes over time with similar resolution compared to that of conventional histology.
Polarization gating is a popular and widely used technique in biomedical optics to sense superficial tissues (colinear detection), deeper volumes (crosslinear detection), and also selectively probe subsuperficial volumes (using elliptically polarized light). As opposed to the conventional linearly polarized illumination, we propose a new protocol of polarization gating that combines coelliptical and counter-elliptical measurements to selectively enhance the contrast of the images. This new method of eliminating multiple-scattered components from the images shows that it is possible to retrieve a greater signal and a better contrast for subsurface structures. In vivo experiments were performed on skin abnormalities of volunteers to confirm the results of the subtraction method and access subsurface information.
The actual skin colorimeters analyse reflect values from a limited number of broad spectral bands and consequently present limited reproducibility and specificity when measuring skin colour. Here, Antera 3D(®) , a new device which uses reflectance mapping of seven different light wavelengths spanning the entire visible spectrum, has been compared with Mexameter(®) MX-18, an established narrow-band reflectance spectrophotometer and with Colorimeter(®) CL-400, an established tristimulus colorimetric instrument. Thirty volunteers were exposed to a controlled ultra-violet B light. Measurements with Antera 3D(®) , Mexameter(®) MX-18 and Colorimeter(®) CL-400 were done before treatment and after 2, 7 and 14 days. Antera 3D(®) showed to have a better sensitivity and specificity than Mexameter(®) MX-18 regarding the melanin parameter. A similar sensitivity between Antera 3D(®) and Mexameter(®) MX-18 was found for erythema determination and also for the Commission Internationale de l'Eclairage L*, a* and b* parameters between Antera 3D(®) and Colorimeter(®) CL-400. Good correlations were observed for all the parameters analysed. Repeatability of Mexameter(®) MX-18 and Colorimeter(®) CL-400 values were lower than that of Antera 3D(®) for all the parameters analysed. Antera 3D(®) , such as Mexameter(®) MX-18 and Colorimeter(®) CL-400, are robust, sensitive and precise equipment for the skin colour analysis. © 2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.
Conference Paper
The healing process of chronic wounds is complex, and the complete pathogenesis is not known. Diagnosis is currently based on visual inspection, biopsies and collection of samples from the wound surface. This is often time consuming, expensive and to some extent subjective procedures. Hyperspectral imaging has been shown to be a promising modality for optical diagnostics. The main objective of this study was to identify a suitable technique for reproducible classification of hyperspectral data from a wound and the surrounding tissue. Two statistical classification methods have been tested and compared to the performance of a dermatologist. Hyperspectral images (400-1000 nm) were collected from patients with venous leg ulcers using a pushbroom-scanning camera (VNIR 1600, Norsk Elektro Optikk AS).Wounds were examined regularly over 4 - 6 weeks. The patients were evaluated by a dermatologist at every appointment. One patient has been selected for presentation in this paper (female, age 53 years). The oxygen saturation of the wound area was determined by wavelength ratio metrics. Spectral angle mapping (SAM) and k-means clustering were used for classification. Automatic extraction of endmember spectra was employed to minimize human interaction. A comparison of the methods shows that k-means clustering is the most stable method over time, and shows the best overlap with the dermatologist’s assessment of the wound border. The results are assumed to be affected by the data preprocessing and chosen endmember extraction algorithm. Results indicate that it is possible to develop an automated method for reliable classification of wounds based on hyperspectral data.