Polarization-sensitive hyperspectral imaging of human skin: From system design to clinical validation (Conference Presentation)

We present the development and validation of a new approach for quantitative functional imaging of human skin based on the machine learning technique for the analysis of the hyperspectral skin images. The considered skin parameters include blood volume fraction, blood oxygenation, melanin content, and the epidermal layer thickness. Additionally, the degree of residual polarization of the reflected light has been analyzed. The validity of the approach has been confirmed by the initial preclinical tests with the tissue-mimicking phantoms, functional in-vivo skin tests, and pilot clinical study of type II diabetic patients. The proposed technique has great potential to be implemented in clinical practice for monitoring and diagnosis of chronic skin ulcers and other relevant diseases.