The goal of the cystoscopy study is to use intrinsic fluorescence without dyes (autofluorescence) spectroscopy to diagnose normal mucosa from papillary tumor during cystoscopy. In the past year we are using using new method for data analysis based on principal component analysis (PCA) followed by least-squares regression (LR). For the ten patients taken with 400 nm excitation, a calibration set and a validation set were randomly formed with five patients each. Analysis with PCA and LR gave 100% sensitivity and 100% specificity for the validation set. The univariate analysis for 400 nm excitation data provided sensitivity and specificity of 100% for the validation set. For eleven patients with 370 nm excitation, a calibration set and a validation set were randomly formed with six and five patients respectively. Analysis with PCA and LR gave 90% sensitivity and 93% specificity for the validation set. With univariate anaylsis resulted in 90% sensitivity and 93% spec ificity for the validation. Normalizing to the patients' normal tissue average improved the results for 370 nm excitation but not significantly for 400 nm excitation. Therefore, papillary tumor can be diagnosed from normal mucosa in a randomly divided set. PCA has the advantage of determining mathematically what the important parts of the spectra are while the univariate method used required user intervention and selecting 460 nm, 600, and 680 nm emissions.
Showing the most recent 10 out of 178 publications
