Simulating the electric field induced in deep brain structures by non-invasive stimulation

Details

Supervisors

Derosiere, Gerard ; Crevecoeur, Frédéric

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil biomédical, à finalité spécialisée

Abstract

enTranscranial electrical stimulation (tES) could hold promise for modulating deep brain structures, yet reproducibility remains a major challenge. Existing approaches often rely on standardized anatomical templates such as the MNI (Montreal Neurological Institute) brain, which overlook individual variability and may misrepresent electric field distributions. This thesis addresses the replicability crisis in tES by investigating how anatomical specificity influences targeting accuracy, with a focus on the orbitofrontal cortex (OFC) and the ventral tegmental area (VTA), two regions implicated in the reward system. The first objective was to quantitatively assess electric field distributions using a high-definition tES montage applied to a large dataset of 50 individual MRIs. A fully automated pipeline was developed to generate individualized simulations, correct segmentation errors affecting midbrain structures, and extract region-specific electric field metrics. The second objective was to develop IndiE-Fusion, a novel approach to visualize group-averaged individualized simulations, combining brain alignment techniques (via Demons registration) and iterative averaging to construct a 'mean brain' representation of the electric field distributions.