Description
Date depot: 8 juillet 2021
Titre: Segmentation, classification and generative models for computer-aided diagnosis of neurological diseases from neuroimaging data
Directeur de thèse:
Olivier COLLIOT (ICM)
Directeur de thèse:
Didier DORMONT (ICM)
Domaine scientifique: Sciences et technologies de l'information et de la communication
Thématique CNRS : Intelligence artificielle
Resumé: Neurological diseases are a major public health problem. Early and accurate diagnosis is essential to provide adequate patient care and design effective clinical trials to find new treatments. Neuroimaging plays a major role in the diagnosis of these disorders. Deep learning techniques offer significant promise to assist diagnosis. However, the design of efficient method is hampered by the limited number of annotated data in this field.
The objective of this project is to design and validate deep learning methods for computer-assisted diagnosis of neurological disorders, and in particular methods that can deal with applications where annotated data is limited. More specifically, three research axes are proposed. The first one concerns the segmentation of brain structures and lesions. The second concerns the development of methods for automatic classification of patients in order to improve differential diagnosis. In both cases, we will explore self-supervised and weakly supervised methods, in particular, contrastive learning. The third research axis aims at designing models that could automatically generate medical reports from imaging data.
The developed approaches will be applied to different patient databases acquired by our partners from the Brain Institute and the Pitié-Salpêtrière Hospital and which concern Parkinson's disease and related syndromes, multiple sclerosis and Alzheimer's disease.
Doctorant.e: Fu Guanghui