Multiscale Simulation of Cortical and Hippocampal Dynamics Using Region-Specific Brain Network Models

Recent advances in computational neuroscience underscore the value of biologically accurate models integrating multimodal neuroimaging to simulate brain activity realistically. Deep brain areas like the hippocampus, crucial for memory and cognition, pose challenges for non-invasive study due to their complex dynamics. The Region-Specific Brain Network Model (RSBNM) addresses this by combining Neural Mass Models, high-resolution BigBrain-derived surface meshes, and DWI-based connectivity to simulate hippocampal activity, achieving approximately 0.8 correlation with empirical EEG, MEG, and BOLD data. This project expands RSBNM to include both cortical and deep structures, employing unified high-resolution meshes for anatomically precise, whole-brain simulations. Implemented within The Virtual Brain (TVB) platform, the enhanced framework leverages detailed structural data from EBRAINS and BigBrain, integrating these inputs into a graph-based neural modeling pipeline. Collaboration with Dr. Viktor Jirsa’s group, specifically Paul Triebkorn and Borana Dollomaja, will utilize their expertise in connectome-based modeling, TVB simulations, neural mass modeling, and high-performance computing. Their methodologies and datasets will facilitate rigorous model development, scalable simulations, and empirical validation, thereby enabling RSBNM to provide detailed insights into brain-wide neural interactions with high anatomical fidelity.

Faculty Supervisor:

Alan Evans

Student:

Partner:

Aix-Marseille Université

Discipline:

Computer science

Sector:

Education

University:

McGill University

Program: