Abstract

Light-based devices, like pulse oximeters and imaging systems, detect light after it interacts with skin. Melanin’s strong optical absorption reduces how much light is detected, and this may cause disparate device performance between lightly and darkly pigmented people. It’s critical that devices work equitably across the full spectrum of pigmentation, but there in an unmet need for a means of device testing where pigment varies while other physiologic variables are constant. We address this need by validating devices in Hampshire swine that have large patches of pigmented and nonpigmented skin (PS, NPS). We characterize n=5 swine by colorimetry and histopathology. Like human skin, swine PS patches show dense epidermal basal melanin (by Fontana Masson) and low individual typology angles(ITAPS° = -67 to -5) while NPS patches show no melanin and high ITA(ITANPS° = 31 to 76, p=0.001 ITAPS vs. ITANPS). Placing duplicate devices on PS and NPS patches in the same animal during validation studies directly compares the impact of pigmentation on device performance while controlling for other physiologic factors. With this split-pigment model we compared clinical pulse oximeters, tissue oximeters, and photoacoustic oximeters placed on PS and NPS during desaturation studies and tissue flap surgery, compared fluorescence intensity (detected with intraoperative fluorescence imaging) between PS and NPS during indocyanine green lymphatic mapping, and also compared the agreement of different colorimeters on skin color assessment in PS and NPS. This model is a novel approach to study how pigment impacts light-based medical modalities, which is critical to ensuring equitable device performance across the full spectrum of skin pigmentation.

Abstract

Citation