Intraoperative identification of brain tumour tissue using Raman spectroscopy

A novel technique in real-time perioperative identification of resection margins of tumour tissue, Raman spectroscopy uses spectral analysis to identify the biochemical composition of tissue (fingerprint). Raman spectra are unique for each tissue and allow for discerning between normal and tumour brain tissue in a minimal volume of sample corresponding to the tip of a microprobe. It is an optical analysis which has not negative impact on the sample studied. The aim of the project is to examine and characterise the spectra of a set of patients with glial tumours. The spectral database will then serve as a source for machine learning methods for identification of tissue types. First, ex vivo samples will be analysed, later also in vivo during surgical removal of tumours. The added value of the project is a larger patient population than in recent studies and also the aim of classifying high- and low-grade gliomas which has a direct impact on treatment strategy. The project relies on an experienced multidisciplinary team.