Titre: Graph Neural Networks for improved EEG analysis in the context of Alzheimer’s disease

Abstract: Alzheimer’s disease is the most common neurodegenerative dementia. It affects 50 million people worldwide and its prevalence is rapidly increasing. Current means of detection are quite expensive and/or invasive. Electroencephalography (EEG), on the other hand, is inexpensive, fairly simple to perform and non-invasive. Recent work by our team has highlighted the interest of using graphs to analyse these EEG data. The objective of the thesis is to study the interest of Deep Learning methods adapted to graphs for the detection of Alzheimer’s patients and EEG markers.