IN VIVO DETECTION OF GLIOBLASTOMA THROUGH MULTIMODAL FIBRE-PROBE SPECTROSCOPY

Glioblastoma (GBM) is the most common and aggressive malignant brain tumour in adults, and the survival rate of patients affected by this disease is strongly dependent on the successful resection of the tumour. In this framework, multimodal optical spectroscopy could provide a fast and label-free tool for improving tumour detection and guiding the surgical removal of diseased tissues. In this study, we used an optical fibre-probe system combining multiple spectroscopic techniques for in vivo examination of normal and GBM tissues in mouse brain. Spectroscopic measurements based on fluorescence, Raman, and diffuse reflectance spectroscopy were performed on anesthetized animals through two optical windows implanted on the head of each mouse. Then, the recorded data were analysed using Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) for obtaining an automated classification of the examined tissues based on the intrinsic spectral information provided by Raman and reflectance spectroscopy. These techniques provided 77% and 97% classification accuracy, respectively, by taking advantage of the intrinsic molecular content of the examined tissues. In particular, the high sensitivity and specificity achieved by means of reflectance spectroscopy indicate that such technique is the most suited for in vivo detection of GBM. The presented results demonstrate the potential of our method for improving the diagnosis of suspicious brain areas during surgery through a very fast spectroscopic inspection, thus helping the surgeon in removing all tumour tissues and reducing the probability of GBM recurrence.