Raman spectroscopy was applied to distinguish the spectroscopic information between normal
cervical tissues (14) and cervical neoplasia (17), including low grade squamous intraepithelial lesions
(6) and high grade squamous intraepithelial lesions (11). Standard pathological sections of these
cervical tissues were measured from superficial to stroma layers. We have normalized significant
Raman peaks, 1250 and 1579-1656 cm<sup>-1</sup> by taking a ratio over a stationary Raman at 1004 cm<sup>-1</sup>, and
successfully discriminated between normal and neoplasm cervical tissues.
Monte Carlo fluorescence model has been developed to estimate the autofluorescent spectra associated with the progression of the Exo-Cervical Intraepithelial Neoplasm (CIN). We used double integrating spheres system and a tunable light source system, 380 to 600 nm, to measure the reflection and transmission spectra of a 50 μm thick tissue, and used Inverse Adding-Doubling (IAD) method to estimate the absorption (μa) and scattering (μs) coefficients. Human cervical tissue samples were sliced vertically (longitudinal) by the frozen section method. The results show that the absorption and scattering coefficients of cervical neoplasia are 2~3 times higher than normal tissues. We applied Monte Carlo method to estimate photon distribution and fluorescence emission in the tissue. By combining the intrinsic fluorescence information (collagen, NADH, and FAD), the anatomical information of the epithelium, CIN, stroma layers, and the fluorescence escape function, the autofluorescence spectra of CIN at different development stages were obtained.We have observed that the progression of the CIN results in gradually decreasing of the autofluorescence intensity of collagen peak intensity. In addition, the existence of the CIN layer formeda barrier that blocks the autofluorescence escaping from the stroma layer due to the strong extinction(scattering and absorption) of the CIN layer. To our knowledge, this is the first study measuring the CIN optical properties in the visible range; it also successfully demonstrates the fluorescence model forestimating autofluorescence spectra of cervical tissue associated with the progression of the CIN tissue;this model is very important in assisting the CIN diagnosis and treatment in clinical medicine.
Monte Carlo modeling was developed for simulating the light induced autofluorescence spectra of colon and cervical tissues at different dysplasia grades.These tissues were frozen sliced into 100 um thickness. The transmittance (Ta) and backscatter spectra (Ra) of these tissues were measured using double integral sphere system. The scattering coefficient, μs, and absorption coefficient, μa, were determined by the inverse adding-doubling (IAD) method from the Ta and Ra. spectra. The fluorescence intensity and spectra of these sliced tissues were measured using spectrometer system with cooled 2D CCD array. We simulated light energy distribution in tissue at 330 nm excitation and then convolution with fluorophores intensity and escape function in each tissue layer. The results of the simulation show: (1)fluorescence spectra change with different tissue characteristics, (2)fluorescence intensity decrease with the development of the dyplasia grades and the mucosa thickness, (3)the relative collagen signal decreases, hemoglobin signal increases, and NADH signal increases along with the dyplasia development. The simulated results matched well in vivo measured results. The approach provides an important means for understanding tissue fluorescence spectra’s changes that are very critical for clinic diagnosis.