Skip to Main content Skip to Navigation
Theses

Parcellisation et analyse multi-niveaux de données IRM fonctionnelles. Application à l'étude des réseaux de connectivité cérébrale.

Abstract : Over the last decade, functional MRI has emerged as a widely used tool for mapping functions of the brain. More recently, it has been used for identifying networks of cerebral connectivity that represent the interactions between different brain areas. In this context, a recent strategy is based on a preliminary parcellation of the brain into functional regions, and then identifying functional networks from a measurement of interactions between each area. The first part of this thesis describes a novel approach for parcellation that produces regions that are homogeneous at several levels. These regions are shown to be consistent with the anatomical landmarks of the processed subjects. In the second part, we propose a new family of statistics to identify significant networks of functional connectivity. This approach enables the detection of small, strongly-connected networks as well as larger networks that involve weaker interactions. Finally, within a classification framework, we developed a group-level study, producing networks that synthesize characteristics of functional networks across the population under study.
Complete list of metadatas

Cited literature [111 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00652609
Contributor : Slim Karkar <>
Submitted on : Thursday, December 15, 2011 - 10:24:17 PM
Last modification on : Thursday, December 10, 2020 - 10:54:22 AM
Long-term archiving on: : Friday, November 16, 2012 - 3:45:11 PM

Identifiers

  • HAL Id : tel-00652609, version 1

Citation

Slim Karkar. Parcellisation et analyse multi-niveaux de données IRM fonctionnelles. Application à l'étude des réseaux de connectivité cérébrale.. Traitement du signal et de l'image [eess.SP]. Université de Strasbourg, 2011. Français. ⟨tel-00652609⟩

Share

Metrics

Record views

633

Files downloads

1136