Speech Processing Techniques for Automatic Detection of Disorders in the Human Brain

Natural Language Processing (NLP) is the branch of artificial intelligence that allows computers to interpret and manipulate language. In this project, we apply NLP to spontaneous speech obtained in cognitively normal elderly persons and in patients who are in the early stages of Alzheimer's disease. We investigate to what extent NLP features of speech are related to individual brain properties: regional volume, functional connectivity and amyloid storage. Specifically, we will determine 1. whether differences in connected speech features between older adults reflect differences in functional brain connectivity and structural brain characteristics 2. whether NLP analysis of connected speech allows one to discriminate between healthy aging, preclinical Alzheimer disease and prodromal/early dementia stage of Alzheimer disease The promotor is Rik Vandenberghe, professor of cognitive neurology, and co-promotor Professor Hugo Van hamme (Center for Speech and Image Processing, Department of Electrical Engineering Sciences).