Face skin hydration and barrier function estimation

Abstract

The skin barrier function remains a critical determinant underlying various skin diseases. Transepidermal Water Loss (TEWL) and skin hydration are two indicators found on the surface of the skin and can be used to assess the health of the skin barrier function. However, clinical instruments such as VapoMeter and Corneometer, utilized for measuring TEWL and skin hydration, lack the accessibility and convenience for consumers seeking fast and portable assessments. In this project, a machine learning approach along with various scientific-based feature extractors were explored to estimate the TEWL and skin hydration from facial selfie images. Specifically, Shape-For-Shading (SFS) was proposed to extract the underlying properties such as the reflectance and albedo from RGB images. Leveraging SFS as feature input enables the machine learning model to form a relationship between the image of a patch of face skin, the reflectance and albedo characteristics, and the TEWL and skin hydration level. Through the exploration of this innovative framework, the study aims to contribute to the development of a non-invasive, fast, portable, and convenient for assessing the skin barrier function and hydration status, thus potentially offering insisghts into personalized skincare and dermatological diagnostics.