The applicants proposed to identify, evaluate and improve the clinical role of optical spectroscopy and imaging for cervical pre-cancer detection. To achieve this goal, the following research projects will be conducted. First, a multi-pixel fluorescence imaging system that has been developed will be evaluated as a post-screening/pre-diagnostic filter for cervical pre-cancer detection. The second goal will be to incorporate diffuse reflectance spectra (measured using a flying spot scanner) from cervical tissues in vivo into our fluorescence based algorithms to determine if it can enhance their sensitivity and specificity for diagnostic applications. The specificity of the current fluorescence based clinical algorithms is limited in part by the presence of inflammatory tissues which appear to be diseased, spectroscopically. These algorithms are based on fluorescence spectra at 337, 380 and 460 nm excitation, which may not be optimum for discriminating between neoplastic and inflammatory tissues. The third objective will be to measure and compare the fluorescence spectra of cells that are predominant in inflammatory tissues to that of cultured neoplastic cervical epithelial cells at a series of excitation wavelengths over the ultraviolet and visible spectral regions. The applicants proposed to identify if there are optimal excitation wavelengths that are specific to inflammatory cells, and we will test these parameters in a future clinical investigation. The final component of our proposed research will be to explore the biochemical and morphological basis of the cervical tissue spectra measured in vivo. A Monte Carlo model of tissue fluorescence will be developed which will utilize the fluorescence spectra from cervical tissue microstructures and will also incorporate effects due to both absorption and scattering to describe the bulk tissue fluorescence.
