Abstract
Pulse oximetry enables real-time, noninvasive monitoring of arterial blood oxygen levels. However, results can vary with skin color, thus detecting disparities during clinical validation studies requires an accurate measure of skin pigmentation. Recent clinical studies have used subjective methods such as self-reported color, race/ethnicity to categorize skin. Melanometers based on optical reflectance may offer a more effective, objective approach to assess pigmentation. Here, we review melanometry approaches and assess evidence supporting their use as clinical research tools. We compare performance data, including repeatability, robustness to confounders, and compare devices to each other, to subjective methods, and high-quality references. Finally, we propose best practices for evaluating melanometers and discuss alternate optical approaches that may improve accuracy. Whilst evidence indicates that melanometers can provide superior performance to subjective approaches, we encourage additional research and standardization efforts, as these are needed to ensure consistent and reliable results in clinical studies.
Similar content being viewed by others
Introduction
An important indicator of health is the percentage of hemoglobin molecules that are bound to oxygen in arterial blood (SaO2). Pulse oximetry is a commonly used medical technology that generates a value (SpO2) which is an estimate of SaO2 based on noninvasive optical measurements. SpO2 measurements are used to inform healthcare decisions in a broad range of clinical settings. However, evidence of SpO2 measurement disparities correlated with biological factors such as skin color have been documented1,2. The significance of these disparities was highlighted during the COVID-19 pandemic when it was found that Black patients were three times more likely than white patients to be diagnosed as having normal SpO2 (normoxemic SpO2) despite having a low SaO2 level (occult hypoxemia)3,4. This was found to have impacts on health care, including delayed detection of COVID-19 infection5.
As part of the response to the impact of these clinically significant racial disparities highlighted during the COVID-19 pandemic, the U.S. Food and Drug Administration (FDA) released a safety communication raising concerns about the accuracy of pulse oximetry measurements across racial groups, and highlighting the importance of understanding and resolving this problem6. In 2022, FDA held a panel meeting to gather input from experts7. While the mechanisms underlying racial disparities in pulse oximeter performance remain unclear, panel members recognized a need for accurate assessments of skin pigmentation in clinical studies, including the potential use of objective approaches8.
This review provides an introduction to melanometry as a potential tool for objective assessment of skin pigmentation, to facilitate assessments of the accuracy of pulse oximetry and other clinical optical sensing devices. Melanometers are devices that use reflected light at visible and near-infrared wavelengths to provide objective, quantitative measurements of pigmentation. Prior reviews of melanometry compared different approaches9,10,11,12,13 or focused on clinical feasibility for specific dermatology applications such as monitoring anti-scarring therapies14,15. Here we quantitatively analyze melanometer performance from published data. Specifically, we evaluate performance attributes including comparison to subjective evaluation methods, inter-melanometer agreement, repeatability, and accuracy versus the established best approaches (gold standards). Finally, we discuss the potential for melanometer best practices, standardization, and alternate approaches warranting future study and development.
The importance of accurately measuring melanin
Melanin is a complex polymer produced in various places in the human body, including the skin. It is a dominant biological chromophore (absorber) in the ultraviolet (UV), visible, and near-infrared spectral regions16,17,18. As a consequence, it is a major contributor to differences in skin color (pigmentation). Melanin can impact a variety of biophotonic technologies that require measurement of light that has propagated through skin. Such devices include spectroscopic devices, for example, pulse oximeters, as well as non-pulsatile regional tissue oximeters that are used for measurements in the skin and brain19,20,21. Researchers have used a variety of approaches to categorize patients whilst evaluating the impact of skin pigmentation on the performance of optical devices. Skin has been categorized as light (white), dark (African American), or intermediate (others)1,2 or using classification methods such as the Fitzpatrick skin phototype (FSP)22,23,24,25,26, Munsell color system27,28,29,30 or Massey Skin Color Score31,32. Because these methods do not measure actual melanin content, they have been criticized for being subjective and susceptible to inter-operator variability 33,34. More recent studies of pulse oximeters have used self-identified race3,5,35,36 or subjective evaluation of race/ethnic origin4,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55 to categorize participants. However, given that many people have mixed ethnicities and that skin pigmentation levels can vary within any ethnic group, conflation of ethnicity with skin pigmentation may also produce misleading results and not permit accurate investigation of any oximeter bias33. Some reports have acknowledged limitations of these subjective approaches4,5.
There is growing support from researchers34,53,56 and professional societies57,58 to address the need for improved methods by implementing objective optical methods to quantify melanin content. It is hoped that by measuring skin pigmentation as a continuous variable, rather than grouping skin tones into discrete subjective categories (e.g., light, intermediate, dark), it will be possible to enroll study participants that more accurately represent the range of skin pigmentation levels necessary. Furthermore, accurate melanin content measurements should provide a higher quality metric for statistical analysis of trends or differences in clinical results. Such data may also be useful for optimizing device algorithms and executing numerical models of light-tissue interactions that enhance quantitative understanding of the role of melanin in optical devices59,60,61.
Over the past 20 years, some scientific fields have widely adopted melanometers as research tools. While commercially available, the vast majority of these devices are not cleared or approved by the FDA as medical devices. In dermatology, melanometers have been implemented for studying vitiligo62,63, scar tissue14,64, melasma65, and psoriasis66,67. Anthropologists use melanometers to measure skin pigmentation in individuals of different ancestry68,69,70. By using techniques that have been successfully implemented in these fields, it may be possible to substantially improve the rigor and quality of clinical studies on optical device performance.
Tissue optics of skin pigmentation
Light-tissue interactions determine visualized skin color, as well as the optical signals detected by pulse oximeters and melanometers. While light propagation across heterogeneous material such as skin depends on many factors, the most significant variables are the wavelength-dependent optical properties and spatial distribution of key chromophores, such as melanin and oxygenated/deoxygenated hemoglobin. In addition to absorption, light is scattered by tissue microstructures such as cells and collagen fibers, resulting in high levels of diffuse reflectance at the tissue surface. Detected signals are also influenced by illumination parameters (e.g., spectral distribution, intensity, angular distribution) and light collection characteristics (e.g., contact vs. standoff geometry, acceptance angle). Light delivered to the skin first interacts with the epidermis, an avascular structure containing melanin (Fig. 1a, b)61,71. The deeper dermis layer is highly vascularized but contains no melanin (Fig. 1a, b)61,71. Epidermis and dermis layer thicknesses vary with anatomical site but are typically 50–150 μm and 1–4 mm, respectively.
In a diverse population, variations in melanin content and epidermal microstructure play significant roles in determining visualized skin tone12,72,73,74,75. Melanocytes located in the basal layer of the epidermis produce melanosomes, which are membrane-bound organelles that synthesize and store melanin72,74,76. Epidermal melanin contains multiple pigments including brown or black eumelanin and red or yellow pheomelanin (Fig. 1c)72,75,77. Skin color is affected by the total melanin content as well as the size, number, shape, and packaging of melanosomes, although the number of melanocytes tends to be constant for a given anatomic site regardless of skin pigmentation72,78,79. A progressive variation in melanosome size with ethnic or geographic origin has also been revealed, with melanosomes in Black skin being the largest (1.44 ± 0.67 µm2 × 10–2) followed by Asian skin (1.36 ± 0.15 µm2 × 10–2) and white skin (0.94 ± 0.48 µm2 × 10–2)72,73,75,80. Melanosomes in white skin are distributed as membrane-bound clusters, whereas melanosomes in Black skin tend to be distributed more individually72,73,74 and Asian skin shows a combination of both individual and clustered melanosomes73. Black skin tends to contain more eumelanin72,81 and ~3–6 times more melanin than white skin, whereas Asian skin tends to contain approximately twofold more melanin than white skin72,74,76,82. Histological images of the epidermis presented in prior articles often show considerable differences between people with different ethnicities (Fig. 1d)72. However, ethnic or geographic categories provide only a moderate degree of correlation with epidermal melanin content70, and extensive variations in melanin content exists within these groups, including by country83.
Both melanin and hemoglobin are stronger absorbers for UV and visible wavelengths and weaker for NIR wavelengths61,84, with melanin exhibiting an exponential decrease with wavelength16,17,84,85,86. Eumelanin and pheomelanin exhibit similar absorption spectra (Fig. 1c) and are often considered as a single chromophore in VIS-NIR (visible and NIR) studies21,85. The epidermal tissue absorption can be calculated as follows87:
where, Mf is the mean volume fraction of melanosomes in the epidermis, \({{{{{{\rm{\mu }}}}}}}_{{{{{{\rm{a}}}}}},{{{{{\rm{mel}}}}}}}\) is the absorption coefficient of a typical melanosome, \({{{{{{\rm{\mu }}}}}}}_{{{{{{\rm{a}}}}}},0}\) is the “baseline” absorption coefficient of epidermal tissue without melanin, \({{{{{{\rm{C}}}}}}}_{{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}}\) is concentration of water, \({{{{{{\rm{\mu }}}}}}}_{{{{{{\rm{a}}}}}},{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}}\) is the absorption coefficient of water and λ is wavelength in nm. The following equations can be used to determine μa,mel16,86 and μa,087,88:
The reduced scattering coefficient of melanin also decreases monotonically with wavelength in the visible wavelength range89, and correlations between scattering parameters and surface density of melanin pigments indicate that melanin contributes to the overall scattering properties of skin tissue89,90. Although scattering may play a role in pulse oximetry racial disparities, given the minimal thickness of the epidermis and relatively small magnitude of change in scattering, the impact of this effect is likely not significant. At typical pulse oximeter wavelengths (660 nm and 940 nm), the absorption coefficient of oxygenated and deoxygenated blood at an average hemoglobin concentration of 150 g/L is much lower than that of highly pigmented epidermis (Mf = 0.43, Fig. 2a)16,91,92.
Clinical measurements across FSP I—VI show increasing melanin content decreases visible and near-infrared reflectance, with stronger changes at shorter wavelengths where melanin absorption is greatest (Fig. 2b)61,93,94,95. Some studies have demonstrated dramatic decreases in reflectance from intermediate to highly pigmented skin93,95 whereas others have shown more uniform changes61,94; this inconsistency is likely due, at least in part, to an imperfect correlation between FSP level and melanin content. Increasing melanin concentration also flattens reflectance spectra in the visible range, weakening features attributed to blood absorption troughs61,93,94,95. The attenuation of UV radiation by melanin in highly pigmented skin reduces the level of UVA and UVB transmission through the epidermis by about 70%72,96.
Subjective skin phototype classification methods
Many methods of classifying skin by pigmentation level have been implemented in dermatology, anthropology, and other disciplines. For example, they have been used to evaluate potential for UV photodamage97,98, evaluate scars and burns99, and provide a biomarker for genetic studies69, as well as to study variations in the performance of optical diagnostic devices.
The most ubiquitous subjective classification method in the literature is FSP22,23,24,25,26, which remains in common medical use. This scale was originally developed in 1975100 as the Fitzpatrick-Pathak skin typing system to assess ultraviolet light sensitivity in “persons with white skin” to select correct initial UVA dose for patients undergoing oral methoxsalen photochemotherapy for psoriasis. This was accomplished via a questionnaire regarding the subject’s response to sun exposure, without regard to skin color100. Later, FSPs V and VI were added to include “brown and black-skinned persons”98,101,102,103. This approach later became a tool for describing skin color and ethnicity103,104. It is worth noting that skin color can refer to constitutive pigmentation – an individual’s inherent pigmentation in the absence of solar/ultraviolet exposure, which was originally relevant to FSP – or facultative pigmentation, which accounts for changes due to sun exposure. The current FSP classification denotes six different skin phototypes depending on the individual’s erythema sensitivity and ability to tan (see tanning and skin types for FSP in Supplementary Fig. 1a). While perceived FSP has been determined via the use of skin color charts105,106, there is no established color palette for perceived FSP I-VI categories, and they have not been mapped to a standardized color space. FSP has proven diagnostic and therapeutic value and has been used to predict the risk of photodamage and skin cancer98, assess the clinical benefits and efficacy of cosmetic procedures103,107, and has been adopted in FDA guidelines for evaluating sunscreen products108. FSP has also been used as a skin color classification tool in a clinical study to support FDA approval of an optical device109 and in several pulse oximeter performance evaluation studies22,23,24,25,26.
Von Luschan’s chromatic scale (VLS) is a skin color classification method based on a set of 36 opaque, colored glass tiles used as a visual reference (see VLS scale in Supplementary Fig. 1b)110. VLS was extensively used in anthropological field research in the 1950s69, but the legacy of von Luschan and VLS has been considered controversial111,112,113. Other subjective classification methods adopted in literature include self-reported skin color114,115,116, racial/ethnic classifications1,2,3,4,5,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55, Munsell color chart27,28,29,30, Massey Skin Color Score31,32, Taylor hyperpigmentation scale117, and scar assessment scales14,64,118. An alternative method that has gained attention recently is the Monk Skin Tone (MST) scale, which is defined by 10 tones and is intended to provide a broader spectrum of pigmentation119,120. The MST tones have been mapped to established color spaces, including RGB and Commission Internationale de l’Eclairage (CIE) LAB.
Although subjective methods have been used for categorization of skin pigmentation levels for many years and the aforementioned methods provide a general basis for skin pigment classification, errors can result from observer/user bias, lighting conditions, and skewed self-reporting11,69,121,122,123,124,125. Several studies have demonstrated that FSP shows weak correlation with skin color and that physicians predominantly assign non-white individuals to FSP IV, V and VI based on their ethnicity, which has proven to be unreliable126,127,128. Lack of reliability can also be caused when FSP, an approach developed for light-skinned people, is implemented to study a diverse population129. Perceived FSP uses a relatively coarse categorization for a continuum of skin tones (fair to dark)125 which causes individual categories such as FSP VI to cover a wide range of melanin concentrations (and reflectance values, Fig. 2b). When determined using the original methodology, FSP values do not account for changes in melanin content due to tanning which may impact optical device performance. Moreover, color changes in dark-pigmented skin are generally overlooked by visual assessment11 and do not address potential intra-individual variations in melanin concentration (immediate pigment darkening or delayed tanning)125.
Melanometry principles
Objective, quantitative and observer-independent measures of skin color can be attained by non-invasive devices known as melanometers. Melanometers can be broadly classified based on their acquisition method, processing technique and melanin metrics (Fig. 3a). In this section, we discuss general operating principles as well as the different implemented methodologies for quantifying pigmentation (summarized in Table 1, example devices in Fig. 3b, see also Supplementary Data 1, a version of Table 1 which also includes device parameter settings).
Data acquisition techniques
Melanometers use visible and near-infrared reflectance measurements. One of the most common approaches involves acquiring narrowband reflectance measurements at selected spectral bands (2–3 wavelengths). The selected wavebands typically include green (relatively strong absorption for melanin), and red and near infrared wavelengths (relatively weak absorption for hemoglobin). To our knowledge, there is only one FDA-cleared narrowband melanometer (Skintel), which was packaged as a component of a light-based skin treatment system130. Typically, narrowband melanometers work in direct skin contact, with light emitting diodes (LEDs) illuminating the skin and a silicon photodetector collecting reflected light. Some narrowband devices are calibrated with a provided set of reflecting white and non-reflecting black plates.
A more rigorous and flexible approach to assessing skin pigmentation can be achieved with melanometers using broadband reflectance spectroscopy, which acquire data at many narrow bands, forming a spectrum. This method allows selection of specific wavelengths for data processing algorithms and enables spectral fitting to chromophore absorption signatures, such as oxyhemoglobin (HbO2), deoxyhemoglobin (HHb), melanin, and water. These systems may use tungsten-halogen lamps, pulsed xenon lamps, or broadband white LEDs depending on the wavelength range of interest, and remitted light is usually detected via a charge-coupled device (CCD) array or silicon photodiode array. Although most reflectance spectroscopic melanometers listed in Table 1 are used in direct skin contact, the Antera 3D and SIAscope II are noncontact devices that acquire large-field 2D maps of skin chromophore concentrations93,131. One key benefit of the spectroscopic approach is the flexibility to use spectral data to calculate both a melanin index metric and colorimetric parameters132,133. Typically, reflectance spectroscopic devices are calibrated using black and white references134,135,136 or a diffuse reflectance standard85.
The final type of melanometer discussed is the colorimeter, which quantifies skin color as perceived by the human vision system using the CIE standard observer model. The distributions that represent spectral sensitivity to light have been standardized as mathematical functions, namely 2° and 10° standard observers (Fig. 3 d). The 2° standard observer is typically used with colorimeters and the 10° standard observer is typically used with reflectance spectroscopic melanometers11. The 2° standard observer represents the average human eye’s spectral sensitivity if viewing colors at an armlength distance from a small field of view whereas 10° standard observer represents visual assessment from a larger field of view and provides better correlation to human color vision11. Colorimeters typically deliver white light to the skin, then either apply RGB filters corresponding to known human eye spectral sensitivity or analyze spectral data from reflectance spectroscopy measurements137,138. Calibration of colorimeters is often performed using a white reference or a pair of black and white plates.
Data processing methods and melanin metrics
As illustrated in Table 1, melanometers primarily generate two types of metrics—melanin/erythema indices and CIELAB colorimetry parameters. However, the underlying measurement methods and data processing algorithms and equations can substantially differ between melanometers reporting the same metric.
Narrowband reflectance melanometers typically generate melanin and erythema indices (MI,EI) using red to near-infrared reflectance measurements9 following studies that showed estimated melanin content could be obtained from absorbance spectra A(λ) derived from reflectance R(λ) as A(λ) = -log10(1/R(λ))82,95,139,140. Since melanin absorption is the primary epidermal absorber in the 600–700 nm region (Fig. 3c), the slope of A(λ) in this region can be used to estimate epidermal melanin content139. Initially slope was based on reflectance near 650 nm and 700 nm140. Other studies have used reflectance at 630–700 nm82,95, and 620–720 nm12,17,141,142,143, and 650–700 nm144. Table 1 illustrates that commercial devices such as the Mexameter MX18 and DSM III use similar algorithms. It is also worth noting that one narrowband device (Skintel) used a different approach, based on simulations of light propagation in tissue130. While the names of metrics and wavelengths used in narrowband melanometers may appear similar, the lack of true standardization makes it difficult to perform effective inter-comparison of data collected by different devices.
Several variations on this approach to measuring melanin index have incorporated additional wavelengths. Area under the intensity curve along the 450–615 nm wavelength interval of reflected light has also been used to evaluate pigmentation and classify skin color145,146. Melanin density, derived by subtracting reflectance at 400 nm from 420 nm, is intended to eliminate the confounding effect of hemoglobin absorption which is similar at both wavelengths147,148,149,150. Single wavelength remittance at 390 nm (Fig. 2a) has been used to estimate epidermal melanin concentration given its shallow penetration depth and independence of blood oxygen saturation151. A method that uses a wide range of spectral reflectance data is proposed to be most accurate. Melanin and hemoglobin metrics have been determined from the skin absorption spectrum ( ~ 500–700 nm) via multiple regression analysis, where the absorbance spectrum is assumed to be a linear summation of the absorptions of melanin and hemoglobin according to Beer-Lambert law152,153,154. Diffuse reflectance models where skin is assumed to be a homogeneous semi-infinite turbid media have also been used to determine chromophore concentrations and light scattering properties of the skin85,155.
Colorimetry devices employ reflectance spectroscopy or RGB reflectance approaches to classify skin according to its visual color appearance. Using the CIELAB color space (Fig. 3e)156, reflectance measurements can be converted into colorimetric quantities. The CIELAB system involves a three-dimensional color space consisting of three axes – L*, a* and b*, where L* represents lightness with values from 0 (black) to 100 (white), a* and b* represent the red/green and yellow/blue attributes on the chroma plane, respectively (Fig. 3e). Although applications of tristimulus colorimetric techniques include estimating visual appearance and chemical analysis137, the CIELAB measurements have been found useful for quantifying skin color, with L* and b* representing pigmentation and a* representing erythema levels in a given individual11,132,157,158,159. Objective classification of skin color has also been attained by use of the Individual Typology Angle (ITA), which can be derived from L* and b* as follows83,160,161,162,163,164:
ITA values have been utilized to categorize skin color. Early work on ITA160 was limited to values above 10o, and these values were used to define four primary categories: very light, light, intermediate, and tan. Subsequent studies considered ITA values as low as –90o and added brown and dark categories (Fig. 3f)163. However, the ITA skin color bins for each category are non-uniform in width, with much wider bin size for darker skin compared to lighter skin. To develop ITA-based categories more closely corresponding to differences in epidermal melanin content, a strategy employing uniform bin sizes may be more appropriate. Amongst all melanometry metrics, ITA is the most commonly evaluated, having been validated against many established approaches such as Fontana–Masson staining161, HPLC165,166 and spectrophotometry165,166.
Comparison between melanometry and subjective methods
Many melanometer studies have compared measurements to subjective classification methods such as FSP. While FSP is an imperfect technique, its prevalent use may enable comparison of data from different studies and thus facilitate standardization between different pigmentation evaluation methods or at least provide a basic check on melanometer validity. We compiled data on correlations between objective melanin metrics and FSP in seven devices (Fig. 4)105,167,168,169,170,171,172,173,174,175. Correlation coefficients between melanometer outputs and other subjective classification tools such as VLS scale and observer rated pigmentation scale have been evaluated for Mexameter MX18 (MI, R = 0.90)176, Dermaspectrometer (MI, R = 0.32–0.63)64 and Minolta Chromameter CR-221-R (L*, R = 0.23–0.51; b*, R = 0.24–0.48)64. Custom derived melanin metrics extracted from reflectance spectroscopic data have also been evaluated against FSP classification scale (R = 0.76 – 0.91)61,145,146, visual subjective grading of pigmentation (R = 0.92)95, and self-reported skin color (R = 0.113)115. Melanometers have shown high inter-operator reliability64,105 compared to subjective classification tools which are often affected by perceptual errors105,170,174,177 and demonstrate poor agreement between observers64.
Good correlation is generally seen between device outputs and FSP, although only approximately 28% of studies showed |R | > 0.75 (Fig. 4). Unlike MI, L* and ITA were inversely proportional to skin pigmentation levels, which is expected based on their definitions in section 4. b* was found to not show statistically significant correlation with FSP (Fig. 4a), which may be due to poor sensitivity to melanin in some pigmentation groups159. Although the moderate to good correlation results (Fig. 4a) indicate device agreement with FSP, the results were highly variable ( | R | = 0.23–0.90).
Different methods have been employed to determine FSP (marker colors in Fig. 4a). FSP was determined using questionnaires in 8/11 studies (20/26 correlation results), using visual perception in 1/11 studies (1/26 correlation results), and using both visual perception and questionnaires in 2/11 studies (5/26 correlation results). While the lone study using visual perception of FSP alone reported strong correlation (R = 0.886, p < 0.001) between FSP and melanometer output105, further comparative analysis between different FSP methods could not be performed due to limited studies. No significant variation in correlation results with respect to the heterogeneity of study population FSP types were identified in the compiled dataset. Some studies demonstrate an unconventional method of determining FSP/skin color from melanometer outputs105,178 (Fig. 4d). Results illustrate moderate discrepancies in MI value ranges for each FSP level, with values from one group generally showing lower MI values for FSP II-IV.
Comparison between melanometers
Given the wide variation in melanometer approaches discussed earlier, there is a need to investigate measurement agreement between different devices to understand device validity and soundness of the technology13. To this end, we compiled data on reported correlations between different melanometer outputs (Fig. 5a, b)68,131,148,178,179,180,181,182,183,184.
Melanometry outputs were generally well correlated between different devices and metrics, with 76% (13/17 results) with |R | > 0.75 (Fig. 5a). Mexameter (MX16, MX18)131,148,178,181,182,184, Dermaspectrometer (DSM)68,179,180,181 and Minolta Chromameter (CR200, CR300)179,181,183,185 were the most frequently correlated devices with other commercial melanometer outputs. Even metrics from different measurement techniques (i.e., colorimetry and narrowband reflectance) were well correlated ( | R | = 0.56 – 0.98). However, a few studies have suggested that MI is a better predictor of melanin content than L*68,179, particularly due to decreased correlation observed between MI and L* in more vascularized body sites of low pigmentation skin with high intra-individual variability (e.g., forehead, |R | = 0.93 vs inner arm, |R | = 0.98)68. This greater sensitivity to blood may be due to using shorter wavelengths—where blood absorption is high—to measure L*. Several custom melanin metrics derived from reflectance spectroscopy data have also been correlated against commercial melanometer outputs ( | R | = 0.78–0.96)60,61,148. However, it should be noted that strong correlation between devices does not necessarily mean that either device actually measures the intended parameter13.
Repeatability and reliability of melanometry
Repeatability is a critical aspect of melanometer performance and may be affected by individual characteristics (age, sex, race, anatomical site, skin surface properties), intra- and inter-individual variability (temporal, physical and mental activity, orthostatic effect, menstrual cycle/menopause), environment conditions (lighting, temperature) and several instrument-related variables9. We compiled data for intra-observer64,105,131,181,183,185,186,187,188,189,190, inter-observer64,167,186,187,191 and inter-instrument186 in vivo repeatability studies where intra-class correlation coefficient (ICC) and/or coefficient of variability (CV) have been reported (Fig. 5c–e). ICCs > 0.90 were considered to indicate excellent reliability, good between 0.75 and 0.90, moderate between 0.50 and 0.75, and poor <0.50192, whereas CVs <10% were considered excellent, good between 10% and 20%, moderate between 20% and 30%, and poor > 30%.
Prior review studies have reported the Minolta Chromameter (CR200/CR300) to be a reliable device for skin color assessment due to its good intra-observer, inter-observer and inter-instrument repeatability compared to other devices9,13. However, these papers included a much smaller compilation dataset of melanometer reliability studies than presented here. The heterogeneity of skin pigmentation of the sampled population varies between different studies. In our compiled repeatability dataset (Fig. 5c–e), 5 out of 13 studies reported the FSP distribution of the study population, of which only 2 studies included participants of FSP VI. For the rest of studies, 5 studies reported race and 3 studies did not report any information on patient skin type or ethno-racial background. No significant trend was observed in the repeatability data for studies that included FSP VI versus studies that did not include FSP VI. As shown in Fig. 5c–e, excellent results (ICC > 0.90 and/or |CV | < 10%) were obtained in approximately 85%, 75%, and 67% of intra-observer, inter-observer, and inter-instrument studies, respectively. It should be noted that most reported repeatability studies tested intra- and inter-observer variations for devices and few inter-instrument repeatability studies have been performed. Only one melanometer measure (b* for DSM II) exhibited poor reliability values (intra-observer, |CV | = 63.9%; inter-observer, |CV | = 32.2%), probably due to the fact that the b* parameter is affected by both pigmentation (i.e., yellowness) as well as skin circulation (likely venous circulation, i.e., blueness) thus leading to greater variability187. However, b* has demonstrated moderate to excellent intra-observer, inter-observer, and inter-instrument reliability in several other studies (Fig. 5c–e). Considering all results, the repeatability of objective melanin metrics via melanometers were generally high, particularly when compared to subjective classification methods for pigmentation (ICC = 0.304-0.87, |CV | = 32.4–50.4%64,167,187).
Cross contamination between pigmentation and erythema in melanometry
Commercial melanometers are intended to quantify melanin and erythema biomarkers, but they generally cannot perfectly isolate the contributions of tissue constituents12. Narrowband melanometers using red and NIR wavelengths are especially susceptible to these effects because their greater optical penetration in tissue increases sensitivity to hemoglobin located in deeper dermal layers (Fig. 2a)12. Due to spectral overlap of hemoglobin and melanin, measurements using a limited number of spectral bands may be subject to crosstalk artifacts15,93,95,151. Methods to separate melanin and hemoglobin signal contributions are typically necessary, yet imperfect. Although tristimulus colorimeter metrics (L*, b*, and ITA) are a measure of perceived skin pigmentation, CIELAB parameters were not intended to extract chromophore-specific information, and changes in melanin or hemoglobin concentration can impact all three indices measured (L* a* b*)12.
The challenging nature of separating melanin and erythema content without interference is readily apparent when evaluating correlation between melanin and erythema metrics (see cross talk between melanin and erythema outputs for a few commercial devices in Supplementary Fig. 2). A high degree of correlation was shown in 85.7% (6 of 7) metric comparisons in commercial melanometers ( | R | > 0.70; Table 2). These results indicate that melanometers may often produce higher erythema readings in darkly pigmented skin without any physiological rationale to support this finding. Studies of Mexameter system measurements during erythema induction by UV irradiation reported a significant decrease in measured pigmentation as cutaneous redness increased, despite actual skin pigmentation being, presumably, constant131,182. Changes in tissue blood volume during orthostasis or application of a pressure cuff have also been shown to affect objective melanin metrics193,194. As most commercial melanometers use contact-based probes to acquire skin reflectance, they include the possibility to exert some pressure on the measurement site which could affect the cutaneous blood content and hence influence the measured color9,95. A few instruments have introduced alternatives to reduce the effects of probe pressure which include incorporation of a plastic mask to distribute the force over a relatively large area (Minolta CM200295), foot-switch model where measurements can be performed without applying any pressure on the measurement site (Mexameter MX16195), and inclusion of elastic spring instrument in the probe that ensures constant pressure application on skin (Mexameter MX18131,196, CL400131). Non-contact melanometers such as the SIAscope II and Antera 3D have overcome such limitations.
Diffuse reflectance spectroscopy (DRS) has been proposed to be a more accurate approach to measure skin chromophore concentrations because it offers rich spectral data and can be combined with light transport models and known chromophore absorption spectra to account for contributions of multiple tissue absorbers12. DRS provides spectral data with a high information content, so it can estimate the concentration of biologically relevant chromophores (e.g., melanin, HbO2, HHb) as well as relatively less abundant chromophores like bilirubin, methemoglobin, and carboxyhemoglobin12. However, DRS has still been reported to be sensitive to cross-contamination effects between melanin and erythema measures60. Individual studies have demonstrated fewer cross-contamination effects for SIAscope II93 and Dermacatch182) melanometers. Single wavelength remittance at 390 nm, which predominantly probes the epidermis, is another approach that has been proposed to predict epidermal melanin concentration with less sensitivity to dermal blood volume151.
Comparison of melanometry to reference approaches
While many papers have compared melanometry measurements to subjective skin tone levels or other melanometer results, the most effective way to validate performance is in comparison to an objective, accurate, and well-established reference187. Validation against non-optical measurements involving direct assessment of tissue samples can provide a scientifically rigorous demonstration of the credibility of melanometers, while also correlating these metrics to biological entities having known physical properties. The primary disadvantage of reference techniques is that they are destructive and require invasive skin biopsy. Perhaps the most promising method for verifying melanometry is to benchmark measurements to melanosome volume fraction16,84.
For decades, tissue melanin content has been measured using histological stain methods. The Fontana-Masson (FM) stain is widely considered the preferred histologic approach for identifying melanin and has been used extensively in both research applications and clinical histopathology (Fig. 6a)161,197,198. Alternative stains such as the Von Kossa (VK) and Warthin-Starry (WS) stains have also been used to quantify epidermal melanin content93,198. Although FM stain is popular, a recent comparative analysis of FM, VK, and WS stains has demonstrated WS stain to provide a more sensitive and specific detection of epidermal melanin compared to FM and VK stains198. Melanin content has been quantified from stained histological slides using different techniques such as mean score density/field using a continuous visual analog scale electronic meter93, surface covered by the staining in the epidermis (Fig. 6a)161,198, intensity of the black stain divided by the area of the epidermis82, and proportion of total area of deep layers epidermis (corresponding to the first two cell layers above the epidermal-dermal junction) of skin sections stained as melanin147. While no standardized approach has been established to quantify epidermal melanin via histological methods and histological markers are subject to errors due to processing, interpretation, and sampling bias, currently they are the best available metric for melanin187,198.
Non-microscopy approaches can also provide high quality results. High-performance liquid chromatography (HPLC) is perhaps the most well-established technique for quantifying melanin markers in tissue samples199,200. A highly sensitive and specific laboratory method, it enables melanin composition to be determined81,165,166,200. Eumelanin consists of 5,6-dihydroxyindole (DHI) and 5,6-dihydroxyindole-2-carboxylic acid (DHICA) moieties, while pheomelanin consists of benzothiazine (BT) and benzothiazole (BZ) moieties. These melanin monomer units can be quantitatively analyzed through specific degradation products by HPLC, including pyrrole-2,3,5-tricarboxylic acid (PTCA) and pyrrole-2,3-dicarboxylic acid (PDCA) for DHICA and DHI moieties of eumelanin, respectively. Pheomelanin moieties BZ and BT can be analyzed as their degradation products, thiazole-2,4,5-tricarboxylic acid (TTCA) and 4-amino-3-hydroxyphenylalanine (4-AHP), respectively. Figure 6b shows an example HPLC chromatogram for dark epidermis. Quantifying moiety contributions from participants in six skin color groups based on ITA classification (Fig. 6c) indicated that proportions of PDCA (35%), PTCA (41%), TTCA (20%), and 4-AHP (4%) are constant regardless of skin pigmentation166. Melanin content can also be estimated spectrophotometrically using optical absorption methods at different wavelengths (e.g., 350, 409, 500, 650 nm) after solubilization of melanin from tissue samples in NaOH or Soluene-350158,161,165,201.
Several studies have attempted to validate commercial melanometers against established measurements (Fig. 6g)93,132,148,157,158,159,161,165,202. Good correlation has been observed between ITA and epidermal melanin content measured via FM staining (R = 0.87, P < 0.0001, Fig. 6d)161, HPLC (R = 0.927, P < 0.0001, Fig. 6e)165 and spectrophotometry (R = 0.922, P < 0.0001, Fig. 6f)165. Correlation between total melanin content obtained by HPLC and spectrophotometric methods has also been assessed, with high correlation (R = 0.98, P < 0.0001161; R = 0.99, P < 0.0001165) observed between the two methods. Findings generally indicated good correlation, with 59% of the results showing |R | > 0.75. This is much better than for subjective skin color classification (28%), which supports the use of melanometers for a more accurate measurement of skin pigmentation. Among the common objective melanin metrics (MI, L* and ITA), 44% of L* and 77% of ITA results showed |R | >0.75, with ITA being the most commonly evaluated metric. It should be noted that despite being a popular metric, MI has only been validated against reference measurements of melanin concentration (histological assessment) in one study, with a modest outcome (R = 0.33, P = 0.02148). It has also been observed that b* alone does not demonstrate a strong correlation with HPLC-determined eumelanin, pheomelanin and total melanin (Table 5 in132). Although melanin was reported as a major contributor to b* values in lighter skin types (R = 0.71, P < 0.00001), this relationship breaks down in darker skin types due to optical masking of yellow light by high concentrations of epidermal melanin159.
A limited number of studies have validated custom melanin metrics from reflectance spectroscopy against established techniques82,147,148. Melanin content extracted from the linear part of the 630–700 nm absorbance spectrum demonstrated strong correlation (R = 0.80, P < 0.00182), whereas melanin density (R420nm - R400nm) demonstrated weak (R = -0.280, P = 0.05148) and moderate (R = 0.68, P < 0.01147) correlation when validated against FM-stained histological melanin content. Despite the benefits of comparing melanometers to established alternative methods, limited data is available, including for many commercial melanometers.
Future directions
Advances in melanometer technology may facilitate efforts to ensure equitable performance in pulse oximeters and other optical devices. It may be possible to improve performance through incremental changes in established reflectance devices or greater changes using innovative emerging technologies. Here, we briefly discuss potential technologies that may address this need.
While we described several reflectance-based techniques above, this is only a subset of proposed methods or possible methods that could be developed in the future for quantifying skin pigmentation. Prior studies have measured reflectance at short visible wavelengths (e.g., 460 nm)125 and calculated the difference between reflectance at 400 and 420 nm147,151. Given the high melanin absorption and small penetration depths offered by short visible wavelength reflectance approaches, it seems probable that such devices would enhance selectivity to the epidermis, thereby minimizing interference from dermal blood. Despite small penetration depths that should mostly avoid interrogating vasculature, hemoglobin absorption is high for short wavelengths (especially the Soret band from ~400 to 450 nm91) and thus may confound melanin measurements. This effect could be mitigated by using short wavelengths for which hemoglobin absorption is also minimized, e.g., 380–400 nm or 450–470 nm151.
Photoacoustic Imaging (PAI) is a technology that exploits the rapid heating of blood vessels under pulsed illumination. PAI delivers nanosecond-duration light pulses to tissue, where rapid absorption generates acoustic pressure waves detectable at the tissue surface by ultrasound transducers. Multispectral PAI can generate maps of tissue chromophore content, although quantitative PAI remains challenging due to complex spectral artifacts203. The potential for PAI to measure melanin concentration in people of varying FSP has been established204. Strong correlation has also been shown between PAI epidermal signal intensity and ITA values measured with a colorimeter (SkinColorCatch) and FSP (although the methodology for determining subject FSP was not described)205. To optimize accuracy of melanin quantification, approaches based on photoacoustic microscopy (PAM)206 capable of sub-melanocyte-scale resolution may provide a more effective solution, and this approach has been investigated for quantifying melanin in ex vivo retinal tissue207. However, PAI and PAM systems are currently too complex and costly for widespread clinical adoption as melanometers.
Another promising approach for melanometry is spatial frequency domain imaging (SFDI)208, which makes use of patterned surface illumination to perform multispectral, depth-selective reflectance imaging and map tissue absorption and scattering coefficients. Recently, SFDI measurements of epidermal absorption coefficients were well correlated with colorimetry measurements from a Chroma Meter CR-400 (Konica Minolta Sensing, Inc., Tokyo, Japan)204. However, absorption coefficients were underestimated for highly pigmented skin, due at least in part to the use of a homogeneous skin model in the algorithm as well as use of a calibration phantom that did not provide optically realistic representation of highly pigmented skin. Multi-layered models can help address some of these shortcomings as they can more accurately isolate layer-specific melanin concentration209. However, more advanced models are required to account for the inhomogeneous distribution of melanin in more heavily pigmented skin89.
Several advanced high resolution optical imaging approaches might also serve as effective melanometers. Confocal microscopy can provide imaging of reflectance or fluorescence at micron-scale resolution and has been used to image melanosomes in the epidermis210,211. Multi-photon imaging to depths of several hundred microns can provide accurate estimation of melanin concentration in vivo, without the need for biopsy 209. Electron paramagnetic resonance (EPR) spectrometry can detect melanin pigments present in skin, hair, and most types of malignant melanomas212. Near-infrared fluorescence has also demonstrated promise as an approach for quantifying epidermal melanin content213. Optical coherence tomography (OCT) represents the standard-of-care for clinical retinal imaging and is capable of achieving the resolution necessary for visualization of epidermal microstructure214. While restricted region of interest for characterization of melanin as well as cost and portability of advanced technologies may be a limiting factor to wide acceptance, progress in these technologies will likely lead to more practical options in the future. Alternately, they may be able to serve as high quality reference approaches to complement or replace histopathology as the recognized best practice for melanometer validation (i.e., gold standard).
Best practices and standardization
Overall, the literature indicates that melanometry devices have the potential to be highly effective tools for pigmentation assessment. However, optimal results will require high quality devices as well as optimized methods for implementation and validation. As with any scientific instrument, it is essential that best practices be developed and widely applied in a consistent manner; thus, there is a critical need to incorporate these practices into international consensus standards.
Consensus guidelines for skin pigmentation measurement using colorimetry and reflectance spectroscopy devices have been published, however one of these documents was developed over 25 years ago (by the European Society of Contact Dermatitis, ESCD)9 and two others focus on cosmetic sun protection testing164,215 and provide similar content. These standards recommend that probes be applied with minimal force to reduce variations caused by displacing blood. This approach aligns with some prior studies140, whereas others have applied firm125 or moderate pressure11. Consensus recommendations also include triplicate measurements with the probe lifted from the skin after each acquisition, a maximum ITA standard deviation of 0.2, and the rapid execution of measurements with the probe steady and perpendicular to the skin. Direct sunlight should be avoided, and ambient temperature should be 19–23 °C. Additionally, measurements should avoid highly heterogeneous regions—such as those containing visible blood vessels, nevi, and scars—as well as regions with discolorations that may be endogenous or exogenous (e.g., tattoos), or those to which cosmetics or medications were applied. Recommended frequency of recalibration—ideally, a process described by the device manufacturer—has varied from before every use11, after every 30 subjects125 or after battery replacement9. In terms of requirements for the melanometer itself, ISO standards164 recommend use of systems with a field of view of at least 8 mm and a lamp with color temperature of 6500 K. CIE colorimetry standards developed for non-biological use also provide recommendations on device specifications156.
Establishing best practices for evaluating skin pigmentation during clinical studies is critical to enable reliable assessment of disparities. The University of California San Francisco (UCSF) Open Oximetry project includes a clinical trial evaluating pulse oximeter performance in a balanced diverse population where skin pigmentation is quantified using multiple color scales as well as colorimeters216. To directly address the optical effects of melanin on pulse oximetry, melanometry should be performed at tissue sites where the oximetry sensor is applied. In prior studies, data have often been collected at sites where melanin concentration is potentially high, such as the arm60,125 head217 or torso218. However, pulse oximeters are commonly used on regions of the index finger (palmar surface and nail) where the range of pigmentation levels is relatively small204. Studies with the Mexameter196 and Chroma Meter CR-400204 provided quantitative evidence that the palmar hand and finger have lower pigmentation than the ventral arm (e.g., melanin index of 42 vs. 240; L* of 60 vs. 39). Furthermore, values measured in the palm of otherwise highly pigmented people were only slightly different than those from people with low pigmentation. Melanocyte content in the nail bed has been measured as 5% that of normal skin and these cells typically do not produce melanin219. The ability to reliably differentiate between low levels of pigmentation in these distal finger sites may then be critical to evaluating impact. While the precision of most melanometers may be sufficient to measure skin sites exhibiting a large melanin content range, including skin proximal to the nail bed, accurately measuring low-pigmentation sites on the finger probably requires greater precision. Data acquired from a commercial spectrometer system may be capable of accomplishing this task effectively [personal communication with Dr. L. Shmuylovich; March 18, 2024]. It is not currently clear whether measurements at higher melanin content sites (e.g., dorsal arm) could serve as a viable surrogate for finger sites. Another practical issue for measuring finger sites includes the challenge of measuring strongly curved surfaces or small regions with bulky instruments and devices that have a large field of view.
It is worth noting that this discussion of measurement site is largely predicated on a racial disparity mechanism involving optical absorption by melanin. However, Monk et al. indicate that perceived colorism, or human perception of skin color by oneself and/or others220 is linked to disparities in health (e.g., blood pressure) that may in turn impact device accuracy 221. Other researchers have noted that skin color is a phenotypical trait that may be associated with other traits having physiological implications such as variations in vascular response222 or hemoglobinopathy 223. To directly test these mechanisms as sources of device performance disparity, measurement sites such as the face may be more appropriate, possibly in combination with racial category.
Melanometer calibration and validation represent another critical topic covered by the ESCD report9. Recommended calibration targets are typically very high and low diffuse reflection, i.e. white and black targets. Numerous articles have noted the use of such targets, particularly with commercial melanometers93,131,181. Fullerton et al. describe the use of white, pink and red plates for repeatability testing9. The ESCD report briefly mentions a nine-color calibration plate for colorimeter variability testing. Other studies have implemented sets of colored tiles (e.g., X-Rite Color Checker with 14 tiles)181,182,224 for performance validation. When using color tiles for colorimeter testing, a color difference approach:
where Δ denotes the difference between reference and test devices) is recommended156. The ESCD report also recommends testing repeatability by measuring a white calibration plate 30 times in 10 second intervals to ensure that the standard deviation of the color difference (ΔE*) is less than 0.07225. Although targets are commonly used for colorimeter assessment, they are probably insufficient for rigorous validation of colorimeters or spectroscopy devices intended for skin measurements. Research has indicated that color tile evaluations overestimate colorimeter accuracy in skin181,182. Additionally, color targets are essentially homogeneous, surface reflecting targets, whereas skin is a multi-layer turbid media into which light may penetrate up to a depth of centimeters. Even colorimeters that show strong inter-device agreement using targets might produce different ITA values in skin if they have different illumination geometries (wide-field imaging vs. point measurement), due to differences in sensitivity to tissue layers. Thus, the use of standardized multi-layered tissue phantoms143 could facilitate standardization of diverse systems. Such phantoms would likely have to incorporate epidermis-simulating layers with a wide range of pigmentation levels21.
Colorimeters and reflectance spectroscopic melanometers can be operated in different settings. CIE has reported recommendations on the use of standard illuminants, standard colorimetric observers as well as illuminating and viewing conditions for basic colorimetry 226. It should be noted that device settings such as illuminants, standard observer, specular component inclusion/exclusion, and measurement geometry can impact the device output and the estimated melanin content. Therefore, ideally the same device settings should be used to compare values obtained with different colorimetric and spectrophotometric instruments11.
Ideally, melanometers would report a standard metric with direct optical and biological relevance. Since ITA is based on the CIELAB color space it is often considered is a standardized metric, yet prior validation of colorimeters has relied on color charts rather than realistic turbid media. Additionally, ITA and MI do not have a direct biological meaning. A more broadly useful metric may be a form of the melanosome volume fraction (Mf) parameter used by Jacques17 to calculate epidermis optical properties. Since melanometers do not measure epidermal thickness or provide a direct measurement of Mf, it may be useful to establish an effective melanosome volume fraction parameter, Mfe, based on an assumed epidermal thickness (e.g., 100 μm). By calibrating ITA or MI outputs to Mfe, direct comparisons between different devices would be possible. A similar calibration of photothermal sensing systems based on histology has been proposed to enable estimation of epidermal melanin mass per volume151.
Concluding remarks
Considerable published evidence supports the use of optical melanometers for assessing skin pigmentation level. Melanometry enables more accurate evaluation of skin pigmentation level than subjective approaches, with a high degree of repeatability and inter-device consistency. Subjective approaches may be sufficient in some clinical studies of optical device robustness to skin pigmentation, yet melanometry will likely enable detection of disparities with greater reliability or with smaller population sizes. Additionally, limited studies indicate strong correlation to objective established measures of epidermal melanin content, which may facilitate more rigorous scientific understanding of clinical results. While devices based on colorimetry and multi-wavelength spectroscopy provide relatively simple and practical tools, higher performance spectroscopy systems appear to enable the greatest accuracy and flexibility. The latter may also be particularly useful for pulse oximetry studies that address the low pigmentation levels in the distal finger.
Moving forward, we see a need for two main research directions. The first is to standardize melanometry through the establishment of best practices for human subject measurements, as well as benchtop and clinical validation of device performance. This should include establishment of approaches to characterize robustness to confounders (e.g., hemoglobin) shown to degrade accuracy. The second is to advance the accuracy of melanometry through improved instrument design and algorithms for reflectance-based devices, and development of alternate optical sensing technologies. Despite these challenges, optical melanometry has the potential to provide effective tools suitable for evaluating robustness of pulse oximeters and other optical diagnostic devices to skin pigmentation, ensuring medical devices achieve healthcare equity in all patients.
References
Bickler, P. E., Feiner, J. R. & Severinghaus, J. W. Effects of skin pigmentation on pulse oximeter accuracy at low saturation. Anesthesiology 102, 715–719 (2005).
Feiner, J. R., Severinghaus, J. W. & Bickler, P. E. Dark skin decreases the accuracy of pulse oximeters at low oxygen saturation: the effects of oximeter probe type and gender. Anesthesia and analgesia 105, S18–S23 (2007).
Sjoding, M. W., Dickson, R. P., Iwashyna, T. J., Gay, S. E. & Valley, T. S. Racial bias in pulse oximetry measurement (letter). N. Engl. J. Med. 383, 2477–2478 (2020).
Valbuena, V. S. M. et al. Racial bias in pulse oximetry measurement among patients about to undergo extracorporeal membrane oxygenation in 2019-2020: A retrospective cohort study. Chest 161, 971–978 (2022).
Fawzy, A. et al. Racial and ethnic discrepancy in pulse oximetry and delayed identification of treatment eligibility among patients with COVID-19. JAMA Intern. Med. 182, 730–738 (2022).
FDA, Pulse oximeter accuracy and limitations: FDA safety communication. Published 2021 Feb 19, (2021).
FDA, Anesthesiology and respiratory therapy devices panel of the medical devices advisory committee meeting, (2022).
Keller, M. D., Harrison-Smith, B., Patil, C. & Arefin, M. S. Skin colour affects the accuracy of medical oxygen sensors. Nature 610, 49–451 (2022).
Fullerton, A. et al. Guidelines for measurement of skin colour and erythema. A report from the Standardization Group of the European Society of Contact Dermatitis. Contact Dermatitis 35, 1–10 (1996).
Kanellis, V. G. A review of melanin sensor devices. Biophys. Rev. 11, 843–849 (2019).
Ly, B. C. K., Dyer, E. B., Feig, J. L., Chien, A. L. & Del Bino, S. Research techniques made simple: cutaneous colorimetry: A reliable technique for objective skin color measurement. J. Invest. Dermatol 140, 3–12.e11 (2020).
Stamatas, G. N., Zmudzka, B. Z., Kollias, N. & Beer, J. Z. Non-invasive measurements of skin pigmentation in situ. Pigment Cell. Res. 17, 618–626 (2004).
Langeveld, M., van de Lande, L. S., O’Sullivan, E., van der Lei, B. & van Dongen, J. A. Skin measurement devices to assess skin quality: A systematic review on reliability and validity. Skin Res. Technol. 28, 212–224 (2022).
Lee, K. C., Dretzke, J., Grover, L., Logan, A. & Moiemen, N. A systematic review of objective burn scar measurements. Burns Trauma 4, 14 (2016).
M. Jaspers, and P. Moortgat, Objective assessment tools: Physical parameters in scar assessment, in Textbook on Scar Management(Springer, 2020), pp. 149-158.
Jacques, S. L. Optical properties of biological tissues: A review. Phys. Med. Biol. 58, R37–R61 (2013).
Jacques, S. L. Quick analysis optical spectra to quantify epidermal melanin and papillary dermal blood content of skin. J. Biophotonics 8, 309–316 (2015).
Lister, T., Wright, P. A. & Chappell, P. H. Optical properties of human skin. J. Biomed. Opt 17, 090901 (2012).
Bickler, P. E., Feiner, J. R. & Rollins, M. D. Factors affecting the performance of 5 cerebral oximeters during hypoxia in healthy volunteers. Anesthesia Analgesia 117, 813–823 (2013).
Sun, X. et al. Skin pigmentation interferes with the clinical measurement of regional cerebral oxygen saturation. Br. J. Anaesth. 114, 276–280 (2015).
Afshari, A. et al. Evaluation of the robustness of cerebral oximetry to variations in skin pigmentation using a tissue-simulating phantom. Biomed. Opt. Expr. 13, 2909–2928 (2022).
Ebmeier, S. et al. A two centre observational study of simultaneous pulse oximetry and arterial oxygen saturation recordings in intensive care unit patients. Anaesth. Intensive Care 46, 297–303 (2018).
Harskamp, R. E. et al. Performance of popular pulse oximeters compared with simultaneous arterial oxygen saturation or clinical-grade pulse oximetry: a cross-sectional validation study in intensive care patients. BMJ Open Respir. Res. 8, e000939 (2021).
Pilcher, J. et al. A multicentre prospective observational study comparing arterial blood gas values to those obtained by pulse oximeters used in adult patients attending Australian and New Zealand hospitals. BMC Pulm. Med. 20, 1–9 (2020).
Ploen, L., Pilcher, J., Beckert, L., Swanney, M. & Beasley, R. An investigation into the bias of pulse oximeters. Respirology 21, 6–6 (2016).
Chan, M. et al. Enabling continuous wearable reflectance pulse oximetry at the sternum. Biosensors 11, 521 (2021).
Adler, J. N., Hughes, L. A., Vivilecchia, R. & Camargo, C. A. Jr Effect of skin pigmentation on pulse oximetry accuracy in the emergency department. Acad. Emerg. Med. 5, 965–970 (1998).
Foglia, E. et al. Accuracy and precision of pulse oximetry in hypoxemic infants. Eur. J. Pediatr 175, 1584–1585 (2016).
Ries, A. L., Prewitt, L. M. & Johnson, J. J. Skin color and ear oximetry. Chest 96, 287–290 (1989).
Foglia, E. E. et al. The effect of skin pigmentation on the accuracy of pulse oximetry in infants with hypoxemia. J. Pediatr. 182, 375–377.e372 (2017).
Harris, B. U. et al. Accuracy of a portable pulse oximeter in monitoring hypoxemic infants with cyanotic heart disease. Cardiol. Young 29, 1025–1029 (2019).
Harris, B. U. et al. Accuracy of pulse oximeters intended for hypoxemic pediatric patients. Pediatr. Crit. Care Med. 17, 315–320 (2016).
Norton, H. Variation in pulse oximetry readings: melanin, not ethnicity, is the appropriate variable to use when investigating bias. Anaesthesia 77, 354–355 (2022).
Okunlola, O. E. et al. Pulse oximeter performance, racial inequity, and the work ahead. Respir. Care 67, 252–257 (2022).
Jubran, A. & Tobin, M. J. Reliability of pulse oximetry in titrating supplemental oxygen therapy in ventilator-dependent patients. Chest 97, 1420–1425 (1990).
Wong, A.-K. I. et al. Analysis of discrepancies between pulse oximetry and arterial oxygen saturation measurements by race and ethnicity and association with organ dysfunction and mortality. JAMA Network Open 4, e2131674–e2131674 (2021).
Wiles, M. D. et al. The effect of patient ethnicity on the accuracy of peripheral pulse oximetry in patients with COVID-19 pneumonitis: a single-centre, retrospective analysis. Anaesthesia 77, 143–152 (2022).
Valbuena, V. S. et al. Racial bias and reproducibility in pulse oximetry among medical and surgical inpatients in general care in the Veterans Health Administration 2013-19: multicenter, retrospective cohort study. BMJ 378, e069775 (2022).
Gottlieb, E. R., Ziegler, J., Morley, K., Rush, B. & Celi, L. A. Assessment of racial and ethnic differences in oxygen supplementation among patients in the intensive care unit. JAMA Intern. Med 182, 849–858 (2022).
Abrams, G. A., Sanders, M. K. & Fallon, M. B. Utility of pulse oximetry in the detection of arterial hypoxemia in liver transplant candidates. Liver Transpl 8, 391–396 (2002).
Avant, M. G., Lowe, N. & Torres, A. Jr Comparison of accuracy and signal consistency of two reusable pulse oximeter probes in critically ill children. Respir. Care 42, 698–704 (1997).
Brooks, J. C. et al. Transcutaneous oxygen saturation accuracy in critically ill children, Research Square https://doi.org/10.21203/rs.2.21938/v1 (2020).
Hinkelbein, J., Koehler, H., Genzwuerker, H. V. & Fiedler, F. Artificial acrylic finger nails may alter pulse oximetry measurement. Resuscitation 74, 75–82 (2007).
Hinkelbein, J., Genzwuerker, H. V., Sogl, R. & Fiedler, F. Effect of nail polish on oxygen saturation determined by pulse oximetry in critically ill patients. Resuscitation 72, 82–91 (2007).
Lee, K., Hui, K., Tan, W. & Lim, T. Factors influencing pulse oximetry as compared to functional arterial saturation in multi-ethnic Singapore. Singapore Med. J. 34, 385–385 (1993).
McGovern, J. P., Sasse, S. A., Stansbury, D. W., Causing, L. A. & Light, R. W. Comparison of oxygen saturation by pulse oximetry and co-oximetry during exercise testing in patients with COPD. Chest 109, 1151–1155 (1996).
Muñoz, X. et al. Accuracy and reliability of pulse oximetry at different arterial carbon dioxide pressure levels. Eur. Respir. J. 32, 1053–1059 (2008).
Ross, P. A., Newth, C. J. & Khemani, R. G. Accuracy of pulse oximetry in children. Pediatrics 133, 22–29 (2014).
Schallom, M., Prentice, D., Sona, C., Arroyo, C. & Mazuski, J. Comparison of nasal and forehead oximetry accuracy and pressure injury in critically ill patients. Heart & Lung 47, 93–99 (2018).
Smyth, R. J. et al. Ear oximetry during combined hypoxia and exercise. J. Appl. Physiol. 60, 716–719 (1986).
Stewart, K. & Rowbottom, S. Inaccuracy of pulse oximetry in patients with severe tricuspid regurgitation. Anaesthesia 46, 668–670 (1991).
Thrush, D. & Hodges, M. R. Accuracy of pulse oximetry during hypoxemia. South. Med. J 87, 518–521 (1994).
Vesoulis, Z., Tims, A., Lodhi, H., Lalos, N. & Whitehead, H. Racial discrepancy in pulse oximeter accuracy in preterm infants. J. Perinatol. 42, 79–85 (2022).
Zeballos, R. J. & Weisman, I. M. Reliability of noninvasive oximetry in black subjects during exercise and hypoxia. Am. Rev. Respir. Dis. 144, 1240–1244 (1991).
Philip, K. E. J. et al. Working accuracy of pulse oximetry in COVID-19 patients stepping down from intensive care: a clinical evaluation. BMJ Open Respir. Res 7, e000778 (2020).
Cabanas, A. M., Fuentes-Guajardo, M., Latorre, K., León, D. & Martín-Escudero, P. Skin pigmentation influence on pulse oximetry accuracy: A systematic review and bibliometric analysis. Sensors 22, 3402 (2022).
Feldman J. APSF statement on pulse oximetry and skin tone, (2021).
ASA, American Society of Anesthesiologists comments on FDA discussion paper: Approach for Improving the Performance Evaluation of Pulse Oximeter Devices Taking Into Consideration Skin Pigmentation, Race and Ethnicity, (2024).
Karsten, A. E., Singh, A., Karsten, P. A. & Braun, M. W. Diffuse reflectance spectroscopy as a tool to measure the absorption coefficient in skin: South African skin phototypes. Photochem. Photobiol. 89, 227–233 (2013).
Wright, C. Y. et al. Diffuse reflectance spectroscopy versus Mexameter((R)) MX18 measurements of melanin and erythema in an African population. Photochem. Photobiol. 92, 632–636 (2016).
Zonios, G., Bykowski, J. & Kollias, N. Skin melanin, hemoglobin, and light scattering properties can be quantitatively assessed in vivo using diffuse reflectance spectroscopy. J. Invest. Dermatol 117, 1452–1457 (2001).
Hegyi, V., Petrovajová, M. & Novotný, M. An objective assessment of melanin in vitiligo skin treated with Balneo PUVA therapy. Skin Res. Technol. 20, 108–115 (2014).
Park, E. S. et al. Application of a pigment measuring device-Mexameter-for the differential diagnosis of vitiligo and nevus depigmentosus. Skin Res. Technol. 12, 298–302 (2006).
Draaijers, L. J. et al. Colour evaluation in scars: tristimulus colorimeter, narrow-band simple reflectance meter or subjective evaluation? Burns 30, 103–107 (2004).
Kanellis, V. G. Objective quantification of melasma severity using melanometers to quantify melanin pigmentation. Biophys. Rev. 12, 1139–1140 (2020).
Pershing, L., Bakhtian, S., Wright, E. & Rallis, T. Differentiation of involved and uninvolved psoriatic skin from healthy skin using noninvasive visual, colorimeter and evaporimeter methods. Skin Res. Technol. 1, 140–144 (1995).
Ahmad Fadzil, M., Ihtatho, D., Mohd Affandi, A. & Hussein, S. Objective assessment of psoriasis erythema for PASI scoring. J. Med. Eng. Technol 33, 516–524 (2009).
Shriver, M. D. & Parra, E. J. 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. 112, 17–27 (2000).
Swiatoniowski, A. K., Quillen, E. E., Shriver, M. D. & Jablonski, N. G. Technical note: comparing von Luschan skin color tiles and modern spectrophotometry for measuring human skin pigmentation. Am. J. Phys. Anthropol. 151, 325–330 (2013).
Norton, H. L. et al. Quantitative assessment of skin, hair, and iris variation in a diverse sample of individuals and associated genetic variation. Am. J. Phys. Anthropol. 160, 570–581 (2016).
Oltulu, P., Ince, B., Kokbudak, N., Findik, S. & Kilinc, F. Measurement of epidermis, dermis, and total skin thicknesses from six different body regions with a new ethical histometric technique. Turk. J. Plast. Surg. 26, 56 (2018).
Brenner, M. & Hearing, V. J. The protective role of melanin against UV damage in human skin. Photochem. Photobiol. 84, 539–549 (2008).
Thong, H. Y., Jee, S. H., Sun, C. C. & Boissy, R. The patterns of melanosome distribution in keratinocytes of human skin as one determining factor of skin colour. Br. J. Dermatol. 149, 498–505 (2003).
Tadokoro, T. et al. UV‐induced DNA damage and melanin content in human skin differing in racial/ethnic origin. FASEB J 17, 1177–1179 (2003).
Alaluf, S. et al. Ethnic variation in melanin content and composition in photoexposed and photoprotected human skin. Pigment Cell Res 15, 112–118 (2002).
Tadokoro, T. et al. Mechanisms of skin tanning in different racial/ethnic groups in response to ultraviolet radiation. J. Invest. Dermatol 124, 1326–1332 (2005).
Thody, A. J. et al. Pheomelanin as well as eumelanin is present in human epidermis. J. Invest. Dermatol 97, 340–344 (1991).
Garcia, R., Mitchell, R., Bloom, J. & Szabo, G. Number of epidermal melanocytes, hair follicles, and sweat ducts in skin of Solomon Islanders. Am. J. Phys. Anthropol. 47, 427–433 (1977).
Szabó, G. in Pigment Cell Biology 99–125 (Academic, 1959).
Szabó, G., Gerald, A. B., Pathak, M. A. & Fitzpatrick, T. B. Racial differences in the fate of melanosomes in human epidermis. Nature 222, 1081–1082 (1969).
Ito, S. & Wakamatsu, K. Quantitative analysis of eumelanin and pheomelanin in humans, mice, and other animals: a comparative review. Pigment Cell Res 16, 523–531 (2003).
Coelho, S. G. et al. Non-invasive diffuse reflectance measurements of cutaneous melanin content can predict human sensitivity to ultraviolet radiation. Exp. Dermatol. 22, 266–271 (2013).
Del Bino, S. & Bernerd, F. Variations in skin colour and the biological consequences of ultraviolet radiation exposure. Br. J. Dermatol. 169, 33–40 (2013).
Jacques, S. L. Origins of tissue optical properties in the UVA, visible and NIR regions, in OSA TOPS on Advances in Optical Imaging and Photon Migration, R. R. Alfano, Fujimoto, J. G., ed. (OSA 1996), pp. 364-371.
Zonios, G. et al. Melanin absorption spectroscopy: new method for noninvasive skin investigation and melanoma detection. J. Biomed. Opt 13, 014017 (2008).
Jacques, S. L. & McAuliffe, D. J. The melanosome: threshold temperature for explosive vaporization and internal absorption coefficient during pulsed laser irradiation. Photochem. Photobiol. 53, 769–775 (1991).
Meglinski, I. V. & Matcher, S. J. Quantitative assessment of skin layers absorption and skin reflectance spectra simulation in the visible and near-infrared spectral regions. Physiol. Meas. 23, 741 (2002).
Saidi, I. S. Transcutaneous optical measurement of hyperbilirubinemia in neonates (Rice University, 1992).
Zhang, X. U. et al. Multidiameter single-fiber reflectance spectroscopy of heavily pigmented skin: modeling the inhomogeneous distribution of melanin. J. Biomed. Opt 24, 127001 (2019).
Bashkatov, A. N. et al. Optical properties of melanin in the skin and skinlike phantoms, in Controlling tissue optical properties: applications in clinical study (SPIE 2000), pp. 219-226.
Prahl, S. Optical absorption of hemoglobin, http://omlc.ogi.edu/spectra/hemoglobin (1999).
Wray, S., Cope, M., Delpy, D. T., Wyatt, J. S. & Reynolds, E. O. R. Characterization of the near infrared absorption spectra of cytochrome aa3 and haemoglobin for the non-invasive monitoring of cerebral oxygenation. Biochim. Biophys. Acta (BBA)-Bioenergetics 933, 184–192 (1988).
Matts, P. J., Dykes, P. J. & Marks, R. The distribution of melanin in skin determined in vivo. Br. J. Dermatol. 156, 620–628 (2007).
Mendenhall, M. J., Nunez, A. S. & Martin, R. K. Human skin detection in the visible and near infrared. Appl. Opt. 54, 10559–10570 (2015).
Stamatas, G. N., Zmudzka, B. Z., Kollias, N. & Beer, J. Z. In vivo measurement of skin erythema and pigmentation: new means of implementation of diffuse reflectance spectroscopy with a commercial instrument. Br. J. Dermatol 159, 683–690 (2008).
Halder, R. M. & Bridgeman‐Shah, S. Skin cancer in African Americans. Cancer 75, 667–673 (1995).
Grandinétti, Vd. S. et al. The thermal impact of phototherapy with concurrent super-pulsed lasers and red and infrared LEDs on human skin. Lasers Med. Sci. 30, 1575–1581 (2015).
Fitzpatrick, T. B. The validity and practicality of sun-reactive skin types I through VI. Arch. Dermatol. 124, 869–871 (1988).
Bae, S. H. & Bae, Y. C. Analysis of frequency of use of different scar assessment scales based on the scar condition and treatment method. Arch. Plast. Surg. 41, 111–115 (2014).
Fitzpatrick, T. B. Soleil et peau. J. Med. Esthet. 2, 33–34 (1975).
Pathak M. A., Jimbow K., Szabo G., & Fitzpatrick T. B. Sunlight and melanin pigmentation, in Photochemical and photobiological reviews(Springer, 1976), pp. 211-239.
Fitzpatrick, T. B. Ultraviolet-induced pigmentary changes: benefits and hazards. Therapeutic Photomed. 15, 25–38 (1986).
Ware, O. R., Dawson, J. E., Shinohara, M. M. & Taylor, S. C. Racial limitations of Fitzpatrick skin type. Cutis 105, 77–80 (2020).
Lin, F. R. et al. Association of skin color, race/ethnicity, and hearing loss among adults in the USA. JARO 13, 109–117 (2012).
Isa, Z. M. et al. The reliability of Fitzpatrick skin type chart comparing to Mexameter (Mx 18) in measuring skin color among first trimester pregnant mothers in Petaling District, Malaysia, Malays. J. Public Health Med 16, 59–65 (2016).
Varughese, P. M. & Krishnan, L. Does color really matter? Reliability of transcutaneous bilirubinometry in different skin-colored babies. Indian J. Pediatr. Dermatol 19, 315 (2018).
Sachdeva, S. Fitzpatrick skin typing: Applications in dermatology. Indian J. Dermatol. Venereol. Leprol 75, 93 (2009).
E. Department of Health, and U. F. Welfare, Sunscreen drug products for over‐the‐counter human drugs; proposed safety, effective and labeling conditions, Federal Register 43, 38206-38269 (1978).
U.S. Food and Drug Administration, Center for Devices and Radiological Health, MelaFind Approval Letter, (2011).
Thomas, N. Hautfarbentafel. by von Luschan. Man 5, 160 (1905).
Smith J. D. WEB Du Bois, Felix von Luschan, and racial reform at the fin de siècle, Amerikastudien/American Studies, 23–38 (2002).
Beatty J. S. On the logic, method and scientific diversity of technical systems: An inquiry into the diagnostic measurement of human skin, (Michigan Technological University, 2017).
Foreman D. WEB Du Bois’s quest to challenge scientific racism, 1906–1932: educating the “City Negro” At the 135th street branch library, (Rutgers University-Graduate School-Newark, 2017).
Reeder, A. I., Hammond, V. A. & Gray, A. R. Questionnaire items to assess skin color and erythemal sensitivity: reliability, validity, and “the dark shift”. Cancer Epidemiol. Biomarkers Prev. 19, 1167–1173 (2010).
Harrison, S. L. & Büttner, P. G. Do all fair-skinned Caucasians consider themselves fair? Prev. Med. 29, 349–354 (1999).
Daniel, L. C., Heckman, C. J., Kloss, J. D. & Manne, S. L. Comparing alternative methods of measuring skin color and damage. Cancer Causes Control 20, 313–321 (2009).
Taylor, S. C., Arsonnaud, S. & Czernielewski, J. The Taylor hyperpigmentation scale: a new visual assessment tool for the evaluation of skin color and pigmentation. Cutis 76, 270 (2005).
Draaijers, L. J. et al. The patient and observer scar assessment scale: A reliable and feasible tool for scar evaluation. Plast. Reconstr. Surg. 113, 1960–1965 (2004).
Monk, Monk skin tone scale, https://skintone.google/get-started.
Schumann, C. et al. Consensus and subjectivity of skin tone annotation for ML fairness. In Proc. Advances in Neural Information Processing Systems 36 (NIPS, 2023).
Colvonen, P. J. Response to: Investigating sources of inaccuracy in wearable optical heart rate sensors. NPJ Digit. Med. 4, 38 (2021).
Bent, B., Enache, O. M., Goldstein, B., Kibbe, W. & Dunn, J. P. Reply: Matters arising ‘Investigating sources of inaccuracy in wearable optical heart rate sensors’. NPJ Digit. Med 4, 39 (2021).
Fider, N. A. & Komarova, N. L. Differences in color categorization manifested by males and females: A quantitative World Color Survey study. Palgrave Commun 5, 1–10 (2019).
Okoji, U. K., Taylor, S. C. & Lipoff, J. B. Equity in skin typing: why it is time to replace the Fitzpatrick scale. Br. J. Dermatol. 185, 198–199 (2021).
Ash, C., Town, G., Bjerring, P. & Webster, S. Evaluation of a novel skin tone meter and the correlation between Fitzpatrick skin type and skin color. Photonics Lasers Med 4, 177–186 (2015).
He, S. Y. et al. Self-reported pigmentary phenotypes and race are significant but incomplete predictors of Fitzpatrick skin phototype in an ethnically diverse population. J. Am. Acad. Dermatol 71, 731–737 (2014).
Halder, R. M. & Nootheti, P. K. Ethnic skin disorders overview. J. Am. Acad. Dermatol 48, S143–S148 (2003).
Taylor, S. C. Skin of color: biology, structure, function, and implications for dermatologic disease. J. Am. Acad. Dermatol 46, S41–S62 (2002).
Pichon, L. C. et al. Measuring skin cancer risk in African Americans: is the Fitzpatrick skin type classification scale culturally sensitive. Ethn. Dis. 20, 174–179 (2010).
Yaroslavsky, I., Childs, J., Altshuler, G. B., Zenzie, H. H. & Cohen, R. Objective measurement device for melanin optical density: dosimetry for laser and IPLs in aesthetic treatments. https://www.bramptonlaserclinic.com/pdf/skintel_technical.pdf (2017).
Matias, A. R., Ferreira, M., Costa, P. & Neto, P. Skin colour, skin redness and melanin biometric measurements: comparison study between Antera((R)) 3D,Mexameter((R)) and Colorimeter((R). Skin Res. Technol 21, 346–362 (2015).
Matsunaka, H., Yamamoto, Y. & Furukawa, F. Non‐invasive quantification of melanin in the stratum corneum: a novel indicator of skin lesions in pigmentation diseases. Skin Res. Technol. 23, 104–111 (2017).
Yun, I. S., Lee, W. J., Rah, D. K., Kim, Y. O. & Park, B. Y. Skin color analysis using a spectrophotometer in Asians. Skin Res. Technol 16, 311–315 (2010).
Mahmoud, B. H. et al. Impact of long-wavelength UVA and visible light on melanocompetent skin. J. Invest. Dermatol 130, 2092–2097 (2010).
Roth, S. Mineralogy, sedimentology and physical properties of two sediment cores from the Southern Tasman Sea (SW Pacific Sector), (1999).
Sun, Y. et al. Deciphering key coloured compounds from sunless tanning reactions. Dyes Pigm 204, 110448 (2022).
Clydesdale, F. M. & Ahmed, E. Colorimetry—methodology and applications. Crit. Rev. Food Sci. Nutr. 10, 243–301 (1978).
Butts, K. Measurement techniques in colorimetry. Quality 54, 18 (2015).
BUCKLEY, W. R. & GRUM, F. Reflection spectrophotometry: use in evaluation of skin pigmentary disturbances. Arch. Dermatol. Res. 83, 249–261 (1961).
Dawson, J. B. et al. A theoretical and experimental study of light absorption and scattering by in vivo skin. Phys. Med. Biol. 25, 695–709 (1980).
Kollias, N. & Baqer, A. Spectroscopic characteristics of human melanin in vivo. J. Invest. Dermatol 85, 38–42 (1985).
Kollias, N. & Baqer, A. On the assessment of melanin in human skin in vivo*,†. Photochem. Photobiol. 43, 49–54 (1986).
Dolotov, L. et al. Design and evaluation of a novel portable erythema‐melanin‐meter. Lasers Surg. Med. 34, 127–135 (2004).
Fajuyigbe, D., Coleman, A., Sarkany, R. P., Young, A. R. & Schmalwieser, A. W. Diffuse reflectance spectroscopy as a reliable means of comparing ultraviolet radiation‐induced erythema in extreme skin colors. Photochem. Photobiol. 94, 1066–1070 (2018).
Eilers, S. et al. Accuracy of self-report in assessing Fitzpatrick skin phototypes I through VI. JAMA Dermatol 149, 1289–1294 (2013).
Pershing, L. K. et al. Reflectance spectrophotometer: The dermatologists’ sphygmomanometer for skin phototyping? J. Invest. Dermatol 128, 1633–1640 (2008).
Dwyer, T., Muller, H. K., Blizzard, L., Ashbolt, R. & Phillips, G. The use of spectrophotometry to estimate melanin density in Caucasians. Cancer Epidemiol. Biomarkers Prev 7, 203–206 (1998).
Wright, C. Y., Lucas, R. M., Kapwata, T., Kunene, Z. & Du Plessis, J. L. Towards a reliable, non-invasive melanin assessment for pigmented skin. Skin. Res. Technol. 25, 100–102 (2019).
Dwyer, T. et al. Cutaneous melanin density of caucasians measured by spectrophotometry and risk of malignant melanoma, basal cell carcinoma, and squamous cell carcinoma of the skin. Am. J. Epidemiol 155, 614–621 (2002).
Pezic, A. et al. Constitutive and relative facultative skin pigmentation among Victorian children including comparison of two visual skin charts for determining constitutive melanin density. Photochem. Photobiol. 89, 714–723 (2013).
Verkruysse, W., Svaasand, L. O., Franco, W. & Nelson, J. S. Remittance at a single wavelength of 390 nm to quantify epidermal melanin concentration. J. Biomed. Opt 14, 014005 (2009).
Hayashi, M. et al. Spectrophotometer is useful for assessing vitiligo and chemical leukoderma severity by quantifying color difference with surrounding normally pigmented skin. Skin Res. Technol. 24, 175–179 (2018).
Masuda, Y., Yamashita, T., Hirao, T. & Takahashi, M. An innovative method to measure skin pigmentation. Skin Res. Technol. 15, 224–229 (2009).
Nishidate, I., Maeda, T., Niizeki, K. & Aizu, Y. Estimation of melanin and hemoglobin using spectral reflectance images reconstructed from a digital RGB image by the Wiener estimation method. Sensors 13, 7902–7915 (2013).
Zonios, G. & Dimou, A. Modeling diffuse reflectance from semi-infinite turbid media: application to the study of skin optical properties. Optics express 14, 8661–8674 (2006).
Z. ISO/CIE 11664-4:, Colorimetry—Part 4: CIE 1976 L* a* b* colour space, (International Organization for Standardization Geneva, Switzerland, 2019).
Hennessy, A. et al. Eumelanin and pheomelanin concentrations in human epidermis before and after UVB irradiation. Pigment Cell Res 18, 220–223 (2005).
Huang, W. S. et al. High correlation between skin color based on CIELAB color space, epidermal melanocyte ratio, and melanocyte melanin content. PeerJ 6, e4815 (2018).
Alaluf, S. et al. The impact of epidermal melanin on objective measurements of human skin colour. Pigment Cell Res 15, 119–126 (2002).
Chardon, A., Cretois, I. & Hourseau, C. Skin colour typology and suntanning pathways. Int. J. Cosmet. Sci. 13, 191–208 (1991).
Del Bino, S. et al. Chemical analysis of constitutive pigmentation of human epidermis reveals constant eumelanin to pheomelanin ratio. Pigment Cell Melanoma Res 28, 707–717 (2015).
Del Bino, S., Duval, C. & Bernerd, F. Clinical and biological characterization of skin pigmentation diversity and its consequences on UV impact. Int. J. Mol. Sci. 19, 2668 (2018).
Del Bino, S., Sok, J., Bessac, E. & Bernerd, F. Relationship between skin response to ultraviolet exposure and skin color type. Pigment Cell Res 19, 606–614 (2006).
ISO, ISO 24444: 2019. Cosmetics—Sun protection test methods—In vivo determination of the sun protection factor (SPF).”
Ito, S., Del Bino, S., Hirobe, T. & Wakamatsu, K. Improved HPLC conditions to determine eumelanin and pheomelanin contents in biological samples using an ion pair reagent. Int. J. Mol. Sci. 21, 5134 (2020).
Del Bino, S., Ito, S., Sok, J. & Wakamatsu, K. 5, 6‐Dihydroxyindole eumelanin content in human skin with varying degrees of constitutive pigmentation. Pigment Cell Melanoma Res 35, 622–626 (2022).
van der Wal, M. et al. Objective color measurements: clinimetric performance of three devices on normal skin and scar tissue. J. Burn. Care. Res 34, e187–e194 (2013).
Sharma, V. K., Gupta, V., Jangid, B. L. & Pathak, M. Modification of the Fitzpatrick system of skin phototype classification for the Indian population, and its correlation with narrowband diffuse reflectance spectrophotometry. Clin. Exp. Dermatol. 43, 274–280 (2018).
Khalid, A. T. et al. Utility of sun-reactive skin typing and melanin index for discerning vitamin D deficiency. Pediatr. Res. 82, 444–451 (2017).
Sommers, M. S. et al. Are the Fitzpatrick skin phototypes valid for cancer risk assessment in a racially and ethnically diverse sample of women? Ethn. Dis. 29, 505 (2019).
Robinson, J. K., Penedo, F. J., Hay, J. L. & Jablonski, N. G. Recognizing Latinos’ range of skin pigment and phototypes to enhance skin cancer prevention. Pigment Cell Melanoma Res 30, 488–492 (2017).
Richard, A., Rohrmann, S. & Quack Lötscher, K. C. Prevalence of vitamin D deficiency and its associations with skin color in pregnant women in the first trimester in a sample from Switzerland. Nutrients 9, 260 (2017).
Young, A. R. et al. Melanin has a small inhibitory effect on cutaneous vitamin D synthesis: A comparison of extreme phenotypes. J. Invest. Dermatol 140, 1418–1426.e1411 (2020).
Linde, K., Wright, C. Y. & Du Plessis, J. L. Subjective and objective skin colour of a farmworker group in the Limpopo Province, South Africa. Skin Res. Technol. 26, 923–931 (2020).
Bailey, S. H. et al. The use of non‐invasive instruments in characterizing human facial and abdominal skin. Lasers Surg. Med. 44, 131–142 (2012).
Treesirichod, A., Chansakulporn, S. & Wattanapan, P. Correlation between skin color evaluation by skin color scale chart and narrowband reflectance spectrophotometer. Indian J. Dermatol. 59, 339–342 (2014).
Damian, D., Halliday, G. & Barnetson, R. S. Prediction of minimal erythema dose with a reflectance melanin meter. Br. J. Dermatol 136, 714–718 (1997).
Wilkes, M., Wright, C. Y., du Plessis, J. L. & Reeder, A. Fitzpatrick skin type, individual typology angle, and melanin index in an African population: steps toward universally applicable skin photosensitivity assessments. JAMA Dermatol 151, 902–903 (2015).
Takiwaki, H., Overgaard, L. & Serup, J. Comparison of narrow-band reflectance spectrophotometric and tristimulus colorimetric measurements of skin color. Skin Pharmacol. Physiol. 7, 217–225 (1994).
Wagner, J. K., Jovel, C., Norton, H. L., Parra, E. J. & Shriver, M. D. Comparing quantitative measures of erythema, pigmentation and skin response using reflectometry. Pigment Cell Res 15, 379–384 (2002).
Clarys, P., Alewaeters, K., Lambrecht, R. & Barel, A. Skin color measurements: comparison between three instruments: the Chromameter®, the DermaSpectrometer® and the Mexameter®. Skin Res. Technol. 6, 230–238 (2000).
Baquie, M. & Kasraee, B. Discrimination between cutaneous pigmentation and erythema: comparison of the skin colorimeters Dermacatch and Mexameter. Skin Res. Technol. 20, 218–227 (2014).
Barel, A. O. et al. The Visi‐Chroma VC‐100®: A new imaging colorimeter for dermatocosmetic research. Skin Res. Technol. 7, 24–31 (2001).
Hua, W., Xie, H., Chen, T. & Li, L. Comparison of two series of non‐invasive instruments used for the skin physiological properties measurements: the ‘Soft Plus’ from Callegari SpA vs. the series of detectors from Courage & Khazaka. Skin Res. Technol. 20, 74–80 (2014).
Uter, W., Benz, M., Mayr, A., Gefeller, O. & Pfahlberg, A. Assessing skin pigmentation in epidemiological studies: The reliability of measurements under different conditions. Skin Res. Technol. 19, 100–106 (2013).
Kerckhove, E., Staes, F., Flour, M., Stappaerts, K. & Boeckx, W. Reproducibility of repeated measurements on healthy skin with Minolta Chromameter CR‐300. Skin Res. Technol. 7, 56–59 (2001).
Lee, K. C. et al. Investigating the intra-and inter-rater reliability of a panel of subjective and objective burn scar measurement tools. Burns 45, 1311–1324 (2019).
Giorgis, A. T. et al. The relationship between melanin and glaucoma: A case-control study. J. Glaucoma 29, 1143–1146 (2020).
Wilhelm, K.-P., Surber, C. & Maibach, H. Quantification of sodium lauryl sulfate irritant dermatitis in man: comparison of four techniques: skin color reflectance, transepidermal water loss, laser Doppler flow measurement and visual scores. Arch. Dermatol. Res. 281, 293–295 (1989).
Nedelec, B., Correa, J. A., Rachelska, G., Armour, A. & LaSalle, L. Quantitative measurement of hypertrophic scar: interrater reliability and concurrent validity. J. Burn Care Res 29, 501–511 (2008).
Gankande, T. et al. Reliability of scar assessments performed with an integrated skin testing device–the DermaLab Combo®. Burns 40, 1521–1529 (2014).
Koo, T. K. & Li, M. Y. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J. Chiropr. Med 15, 155–163 (2016).
Stamatas, G. N. & Kollias, N. Blood stasis contributions to the perception of skin pigmentation. J. Biomed. Opt 9, 315–322 (2004).
Takiwaki, H. & Serup, J. Variation in color and blood flow of the forearm skin during orthostatic maneuver. Skin Pharmacol. Physiol. 7, 226–230 (1994).
Yoshimura, K. et al. Usefulness of a narrow-band reflectance spectrophotometer in evaluating effects of depigmenting treatment. Aesthetic Plast. Surg. 25, 129–133 (2001).
Nedelec, B. et al. Skin characteristics: normative data for elasticity, erythema, melanin, and thickness at 16 different anatomical locations. Skin Res. Technol. 22, 263–275 (2016).
Carriel, V. S. et al. A novel histochemical method for a simultaneous staining of melanin and collagen fibers. J. Histochem. Cytochem 59, 270–277 (2011).
Joly‐Tonetti, N., Wibawa, J. I., Bell, M. & Tobin, D. Melanin fate in the human epidermis: a reassessment of how best to detect and analyse histologically. Exp. Dermatol. 25, 501–504 (2016).
Pena, A.-M. et al. In vivo melanin 3D quantification and z-epidermal distribution by multiphoton FLIM, phasor and Pseudo-FLIM analyses. Sci. Rep. 12, 1–18 (2022).
Alaluf, S. et al. Variation in melanin content and composition in type V and VI photoexposed and photoprotected human skin: the dominant role of DHI. Pigment Cell Res 14, 337–347 (2001).
Wakamatsu, K. & Ito, S. Advanced chemical methods in melanin determination. Pigment Cell Res. 15, 174–183 (2002).
Kongshoj, B., Thorleifsson, A. & Wulf, H. C. Pheomelanin and eumelanin in human skin determined by high‐performance liquid chromatography and its relation to in vivo reflectance measurements. Photodermatol. Photoimmunol. Photomed. 22, 141–147 (2006).
Cox, B. T., Laufer, J. G., Beard, P. C. & Arridge, S. R. Quantitative spectroscopic photoacoustic imaging: A review. J. Biomed. Opt 17, 061202 (2012).
Phan, T. et al. Quantifying the confounding effect of pigmentation on measured skin tissue optical properties: A comparison of colorimetry with spatial frequency domain imaging. J. Biomed. Opt 27, 036002–036002 (2022).
Li, X. et al. Optoacoustic mesoscopy analysis and quantitative estimation of specific imaging metrics in Fitzpatrick skin phototypes II to V. J. Biophotonics 12, e201800442 (2019).
Attia, A. B. E. et al. A review of clinical photoacoustic imaging: Current and future trends. Photoacoustics 16, 100144 (2019).
Shu, X., Li, H., Dong, B., Sun, C. & Zhang, H. F. Quantifying melanin concentration in retinal pigment epithelium using broadband photoacoustic microscopy. Biomed. Opt. Express 8, 2851–2865 (2017).
Cuccia, D. J., Bevilacqua, F., Durkin, A. J. & Tromberg, B. J. Modulated imaging: Quantitative analysis and tomography of turbid media in the spatial-frequency domain. Opt. Lett. 30, 1354–1356 (2005).
Saager, R. B. et al. In vivo measurements of cutaneous melanin across spatial scales: using multiphoton microscopy and spatial frequency domain spectroscopy. J. Biomed. Opt 20, 066005–066005 (2015).
Majdzadeh, A. et al. Real‐time visualization of melanin granules in normal human skin using combined multiphoton and reflectance confocal microscopy. Photodermatol. Photoimmunol. Photomed. 31, 141–148 (2015).
Rajadhyaksha, M., Grossman, M., Esterowitz, D., Webb, R. H. & Anderson, R. R. In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast. J. Invest. Dermatol 104, 946–952 (1995).
Godechal, Q., Ghanem, G. E., Cook, M. G. & Gallez, B. Electron paramagnetic resonance spectrometry and imaging in melanomas: comparison between pigmented and nonpigmented human malignant melanomas. Mol. Imaging 12, 7290–2012 (2013).
Kalia, S. et al. Melanin quantification by in vitro and in vivo analysis of near‐infrared fluorescence. Pigment Cell Melanoma Res 31, 31–38 (2018).
Boone, M. A., Norrenberg, S., Jemec, G. B. & Del Marmol, V. High-definition optical coherence tomography imaging of melanocytic lesions: a pilot study. Arch. Dermatol. Res. 306, 11–26 (2014).
ISO, ISO 24442: 2022. Cosmetics — Sun protection test methods — In vivo determination of sunscreen UVA protection.
Open Oximetry protocol for “skin color quantification for pulse oximeter human study protocol, https://openoximetry.org/study-protocols/.
Howard, J. J., Sirotin, Y. B., Tipton, J. L. & Vemury, A. R. Reliability and validity of image-based and self-reported skin phenotype metrics. IEEE Trans. Biom. Behav Identity Sci 3, 550–560 (2021).
Coelho, S. G., Miller, S. A., Zmudzka, B. Z. & Beer, J. Z. Quantification of UV-induced erythema and pigmentation using computer-assisted digital image evaluation. Photochem. Photobiol. 82, 651–655 (2006).
Gunes, P. & Goktay, F. Melanocytic Lesions of the Nail Unit. Dermatopathology (Basel) 5, 98–107 (2018).
Monk, E. P. Jr Colorism and physical health: evidence from a national survey. J. Health Soc. Behav 62, 37–52 (2021).
Crooks, C. J. et al. Inverse association between blood pressure and pulse oximetry accuracy: an observational study in patients with suspected or confirmed COVID-19 infection. Emerg. Med. J. 40, 216–220 (2023).
Drew, R. C., Charkoudian, N. & Park, J. Neural control of cardiovascular function in black adults: implications for racial differences in autonomic regulation. Am. J. Physiol. Regul. Integr. Comp. Physiol. 318, R234–R244 (2020).
Patterson, S., Sandercock, N. & Verhovsek, M. Understanding pulse oximetry in hematology patients: Hemoglobinopathies, racial differences, and beyond. Am. J. Hematol. 97, 1659–1663 (2022).
Choi, H., Choi, K., & Suk, H.-J. Performance of the 14 skin-colored patches in accurately estimating human skin color, in Electronic Imaging, Computational Imaging XV 2017 (Society for Imaging Sciences and Technology 2017), pp. 62-65.
Chroma-Meter CR-300/CR-310/CR-321/CR-331/CR-331C, Minolta Camera Co., Ltd., Osaka, Japan, 1991, https://www.konicaminolta.com.cn/instruments/download/manual/pdf/CR-300.pdf.
C. CIE, Technical report: colorimetry, Commission Internationale de l'Éclairage Central Bureau. Vienna, Austria. (2004).
Kyriacou, P. A., & Allen, J. Photoplethysmography: technology, signal analysis and applications (Academic Press, 2021).
Cios, A. et al. Effect of different wavelengths of laser irradiation on the skin cells. Int. J. Mol. Sci. 22, 2437 (2021).
Sarna, T., & Swartz, H. The physical properties of melanins, The pigmentary system: physiology and pathophysiology, 311-341 (1988).
Osto, M., Hamzavi, I. H., Lim, H. W. & Kohli, I. Individual typology angle and Fitzpatrick skin phototypes are not equivalent in photodermatology. Photochem. Photobiol. 98, 127–129 (2022).
Acknowledgements
We gratefully acknowledge support for this work from FDA’s Office of Minority Health and Health Equity. We are appreciative of many useful discussions with the IEC/ISO Joint Working Group on Oximetry, Mr. Bob Kopotic (Edwards Lifesciences), Dr. Wim Verkruysse (Philips Research), Mr. Paul Batchelder (Clinimark), Dr. Leo Shmuylovich (Washington Univ. School of Medicine in St. Louis), Prof. Tony Durkin (Beckman Laser Institute, Univ. of California, Irvine), Prof. Yu Chen (Univ. of Massachusetts, Amherst), and Prof. Kung-Bin Sung (Taiwan Univ.). We would also like to acknowledge Dr. Wei-Chung Cheng’s (FDA) expertise in color assessment and contributions to the paper.
Author information
Authors and Affiliations
Contributions
S.V, W.C.V., S.W., and T.J.P. completed the review, extraction, and quality assessment of papers. S.V. and T.J.P. drafted the initial version of the manuscript. W.C.V. and S.W. provided critical intellectual feedback. All authors approved the final version for publication.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Communications Medicine thanks Leo Shmuylovich and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Vasudevan, S., Vogt, W.C., Weininger, S. et al. Melanometry for objective evaluation of skin pigmentation in pulse oximetry studies. Commun Med 4, 138 (2024). https://doi.org/10.1038/s43856-024-00550-7
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s43856-024-00550-7
- Springer Nature Limited