Multimodal & Multiscale Neuroimaging Integration for Cognitive Function-Preserving Brain Tumour Resection – MICBrainPres

The main goal of this project is to harness the latest advances on machine learning-based neuroimage processing technologies to improve function-preserving brain tumour resection. Identifying eloquent brain regions is fundamental to performing tumour resection while preserving a maximum level of cognitive function. Despite the sustained advance in predicting subject-level cognitive abilities from neuroimaging data, current approaches lack sensitivity and specificity in identifying eloquent brain regions. This hinders neuroimaging’s usefulness for pre-surgical planning as a tool to predict the preservation of cognitive function after tumour resection. In this project, we propose that using subject-specific parcellations, derived from functional and diffusion MRI through deep-learning technologies, will achieve the needed sensitivity and specificity to locate eloquent areas pre-surgically and to predict cortical reshaping after tumour resection.