Sélection de variables dans un cadre Bayésien de traitement de données d'IRM fonctionnelle

Christine Bakhous 1, * Florence Forbes 1, * Thomas Vincent 1, 2 Lotfi Chaari 1, * Michel Dojat 3, * Philippe Ciuciu 2
* Corresponding author
1 MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : The General Linear Model (GLM) is an important framework for analyzing event-related functional MRI (fMRI) data. Studies usually assume that all delivered stimuli possibly generate a BOLD response everywhere in the brain although activation is likely to be induced by only some of them in speci c brain areas. To face this issue, we propose within a Joint Detection Estimation (JDE) framework, a variable selection procedure that automatically selects the conditions according to the brain activity they elicit. It follows an improved activation detection that we illustrate on real data.
Document type :
Conference papers
Complete list of metadatas

Contributor : Florence Forbes <>
Submitted on : Wednesday, January 23, 2013 - 11:12:01 AM
Last modification on : Thursday, March 14, 2019 - 7:54:04 AM


  • HAL Id : hal-00780088, version 1



Christine Bakhous, Florence Forbes, Thomas Vincent, Lotfi Chaari, Michel Dojat, et al.. Sélection de variables dans un cadre Bayésien de traitement de données d'IRM fonctionnelle. 44e Journées de Statistique, Société Française de Statistique, May 2012, Bruxelles, Belgique. ⟨hal-00780088⟩



Record views