Raman and fluorescence spectroscopy of in vitro skin tissue for diagnostics and monitoring

Abstract

Raman spectroscopy is a vibrational scattering technique that uses a laser to measure the spectral fingerprint of a molecule. Skin Raman has applications for noninvasive cancer diagnostics and glucose sensing. A major challenge is that skin also emits autofluorescence. This time-varying autofluorescence background reduces the Raman signal to noise ratio (SNR) and correlates spectra taken from a single patient. Before clinical trials, instruments need to be tested and developed on a human skin substitute exhibiting the same challenges. We establish a fully stratified in vitro reconstructed skin tissue model as an optical substitute with similar autofluorescence, photobleaching, and Raman peaks as in vivo human skin. Using this tissue model, we show a new procedure to manipulate tissue glucose, detect using Raman, and confirm using high performance liquid chromatography. It may also be possible to separate Raman and fluorescence in time to improve SNR. But, the near infrared (NIR) autofluorescence lifetime of skin has not been previously characterized. We built a sensitive and fast time-domain electro-optic setup and report the first skin NIR lifetime using the tissue model: [tau] = 118 - 524 ps. This work is the most extensive characterization of NIR autofluorescence of skin thus far. Researchers can use the tissue model to test, compare, and improve their instruments and algorithms for Raman sensing. It can also be used to investigate the mechanism of action for cancer Raman, study hyperspectral autofluorescence lifetime, compare optical designs, and help bring skin Raman to widespread clinical use.