Establishing a causal link between the physiologic range of skin chromophore concentrations and physiologically relevant regions of CIELAB color space

Assessment of skin color by colorimetry can objectively measure skin disease progression and can ensure that clinical trials recruit diverse populations. Colorimeters measure color in 3-dimensional L*a*b* color space (L*=lightness/darkness, a*=green/redness, and b*=blue/yellowness). Studies have empirically found that 1) skin L*, a*, b* values are constrained in L*a*b* space and 2) that the angle between (L*=50, b*=0) and a given subject's L*, b* values in the L*-b* plane (the Individual Typology Angle, ITA) may be a surrogate for skin melanin content. We sought to determine the causal link between skin chromophore distribution/concentration and L*a*b* values and hypothesized that this would 1) explain previous empirically-derived observations regarding human skin in L*a*b* space and 2) inspire novel methods for solving the inverse problem where skin chromophore concentrations are determined from measured L*a*b values. We applied the adding-doubling method to a 3-layer skin model with varying melanosome fraction (M_frac, 0-43%) and blood volume fraction (B_frac, 0.2-7%) to generate a physiologic skin diffuse reflectance spectra that were then converted to L*a*b*. We found that reflectance spectra derived by varying B_frac at constant M_frac generate non-overlapping iso-melanin banana-shaped curves that recapitulated the empirically defined shape and limits of the region of the L*b* plane observed in human studies. Furthermore, iso-melanin curves provided a theoretical basis between ITA and melanin content and demonstrated that ITA is a poor surrogate for melanin when B_frac is low. Overall, this study provides novel insights into how colorimeter-derived L*a*b* values can be used to extract useful information about skin chromophore concentration.