Karametria (statistical analysis of brain structures) – KaraMetria

This project entitled KaraMetria, from Kara (head, brain in ancient Greek) and Metria (measure), aims at developing new algorithms for the statistical analysis of anatomical shapes extracted from brain magnetic resonance images, such as cortical ribbons, white matter fiber bundles, or basal ganglia. Most existing brain morphometry techniques use voxel-based statistics that indirectly reflect the underlying geometrical structure of the shapes under study. Our goal is therefore to devise methods that work directly with the geometrical objects of interest whenever available after a preliminary image segmentation step. As a result, we will provide the neuroscience community with efficient new software tools to study both the normal and abnormal across-subject brain variability, and discover new biological markers specific to certain pathologies. Several databases of patients affected with depressive disorders will be analyzed within the scope of this project using the newly developed methods.