Implementation of autofluorescent excitation-emission matrices for differentiation of cervical tissue

This paper is deals with the implementation of autofluorescent excitation-emission matrices for the differentiation of cervical tissue. Autofluorescence diagnosis is characterized by the stimulation of endogenous fluorophores, such as, collage, NADH, and FAD in normal and abnormal tissue. !t has minimum invasion and it is in real time. In this research, we developed a small autofluorescence spectra measurement system. It consists of a sensitive PMT with a double gratings spectrometer, a Y type fiber including 18-excitation to 19-emission fibers, and xenon lamp with monochromate as multi-wavelength light source. We utilized this autofluorescence spectra and excitation-emission matrices (EEM) in cervical cancer study and measured normal and abnormal tissues (cervical intraepithelial neoplasia, CIN squamous cell carcinoma: SCC). EEM represented multi-excited wavelengths induced by different range wavelength of the emission. It allows us to get more information about tissues' characteristic. We used partial least squares (PLS) to analyze these data of the spectra and EEM. The result shows that the diagnostic ability by autofluorescence technology can be applied to ClN and using PLS analyzed pathological changes of the cervical tissue has a discriminative property. We also got high sensitivity and specificity to distinguish between normal endo and exo cervix by combining EEM and PLS . In the future, we will build autofluorescence EEM database for clinic diagnosis information.