Non-Invasive Estimation of Skin Thickness by near Infrared Diffuse Reflection Spectroscopy—Separate Determination of Epidermis and Dermis Thickness

Abstract

This paper reports a non-invasive method for estimating skin thickness by using near infrared diffuse reflectance (NIR-DR) spectroscopy. Skin thickness is an important skin property in cosmetology, dermatology and pharmaceutical science. It varies significantly between the face and other body parts and changes with age and environment factors. Differences in skin thickness reflect the structural conditions of the epidermis and dermis. However, current methods for measuring skin thickness are complex and require cumbersome equipment. Therefore, we herein propose a NIR-DR and partial least squares (PLS) regression method for non-destructive skin thickness estimation. NIR-DR spectra were measured for UV-irradiated and non-irradiated skin on the backs of hairless mice and denuded back skin. Skin thickness increased gradually with the amount of UV irradiation, while changing little with physiological age. To obtain reference data, optical microscope measurements were carried out for denuded skin and total thickness and thickness of epidermis and dermis were determined. PLS regression was performed for the whole spectrum and selected regions specific for particular functional groups in order to estimate total thickness and thickness of epidermis and dermis. In the 6939–5990 cm−1 and 5242–4609 cm−1 regions, the bands due to amide groups of proteins appeared to contribute to models with the best coefficients of determination ( R2CV) and standard error of cross-validation results; for total thickness, thickness of epidermis, and dermis, these were 0.79 and 25.80 urn, 0.72 and 21.87 μm and 0.77 and 8.16 μm, respectively. The results revealed that the wavenumbers contributing to the prediction of skin thickness differ between the epidermis and dermis.