Équipe IMAGeS : Images, Modélisation, Apprentissage, Géométrie et Statistique

Séminaire du 11/07/2017, 14h00

De Équipe IMAGeS : Images, Modélisation, Apprentissage, Géométrie et Statistique
Aller à la navigation Aller à la recherche

mardi 11 juillet 2017, 14h00

Cerebral Cortex Parcellation Using Quantitative MRI

Conférencier : Mariem Bouhaha

Understanding the amazingly complex human cerebral cortex requires a parcellation mapping its major subdivisions, called cortical areas. Most previous parcellations were based on only one neurobiological property (such as architecture, function, connectivity or topography). However, it would be interesting to quantitatively characterize cortical areas, delineate their borders and document their properties. The first aim behind this work is to make accurate areal maps using multiparametric (or quantitative) MR Imaging. The second objective is to reproduce the submitted parcellation across individual subjects and accurately predict their cortical parcellations based on their individual MRI intensities. In this seminar, I will start with a description of the MRI data in hand and the main preprocessing steps used. The second part will be dedicated to the clustering approach where I will present the algorithms used to make the parcellation. Finally, I will demonstrate some results achieved so far.