Multivariate analysis of endometrial tissue fluorescence spectra

摘要

Background and Objective: The detailed multivariate analysis of endometrial tissue fluorescence spectra was done. Spectra underlying features and classification algorithm were analyzed. An effort has been made to determinate the importance of neopterin component in endometrial premalignization. Study Design/Materials and Methods: Biomedical tissue fluorescence was measured by excitation with the Nd YAG laser third harmonic. Multivariate analysis techniques were used to analyze fluorescence spectra. Biomedical optics group at Vilnius University analyzed the neopterin substance supplied by Institute of Medical Chemistry and Biochemistry of Innsbruck University. Results: Seven statistically significant spectral compounds were found. The classification algorithm classifying samples to histopathological categories was developed and resulted in sensitivity of 80% and specificity 93% for malignant vs. hyperplastic and normal. Conclusions: Fluorescence spectra could be classified with high accuracy. Spectral variation underlying features can be extracted. Neopterin component might play important role in endometrial hyperplasia development