Diagnosing urothelial carcinoma through multiple spectroscopic techniques

摘要

Urothelial carcinoma (UC) is the most common type of bladder cancer, and its treatment depends from both tumourinvasiveness (stage) and aggressiveness (grade). The gold standard for detecting UC is white-light cystoscopy, followedby tissue biopsy and histopathological examination; however, such process is invasive, time-consuming, operatordependentand prone to sampling errors. In this framework, optical spectroscopy techniques can provide a fast, label-freeand non-invasive tool for improving diagnosis. Thus, we combined auto-fluorescence, diffuse reflectance and Ramanspectroscopy in a compact and transportable setup based on an optical fibre-bundle probe. This experimental setup wasused for studying fresh biopsies of urothelial tumour (129 samples) and healthy bladder (40 samples) collected from 78patients undergoing Transurethral Resection of Bladder Tumours (TURBT). The recorded data were analysed usingPrincipal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) for obtaining an automated classificationof the examined samples based on the intrinsic spectral information provided by all three techniques. We found thathealthy and diseased tissues showed significant spectral differences for each technique, resulting in high accuracy (up to90%) from PCA-LDA routines. While fluorescence spectroscopy seems sensitive enough for detecting UC, we foundthat a multimodal approach is crucial for obtaining high discriminating capability (>80%) in grading and staging tumourbiopsies. In conclusion, the presented strategy generates results similar to gold standard histology, but in a fast and labelfreeway, offering the potential for endoscopic in vivo applications.