Development of near-infrared spectroscopy system for skin moisture

Abstract

Water, which is the largest component of the human body, is used as a biomarker to diagnose health conditions and diseases. In particular, skin can be used not only for diseases but also for research and industry of skin care. In order to measure skin moisture, near-infrared spectroscopy using the property that near-infrared ray penetrates the human body well can be used. However, the conventional algorithm for extracting optical properties have limitations in skin measurement. In this study, to solve this problem, high-resolution light intensity information was collected from the skin using a CMOS image sensor. In addition, an artificial neural network model is developed that predicts skin optic properties by replacing conventional algorithms. In order to verify the reliability of the system designed in this study, artificial neural network learning results, solid phantom experiment, and liquid phantom experiment were performed. As a result, the artificial neural network model predicted optical properties better than the conventional algorithm.