Soutenance de la thèse de Qi Wang (Equipe BaNCo )

14 février 2020

Multivariate group analysis for functional neuroimaging : conceptual and experimental advances

Analyses de groupe multivariées pour la neuroimagerie fonctionnelle : avancées conceptuelles et expérimentales

Vendredi 14 février à 14h

Lieu : salle de SEMINAIRE au 2 ème étage (FRUMAM) à St Charles


Title : Multivariate group analysis for functional neuroimaging : conceptual and experimental advances

Titre : Analyses de groupe multivariées pour la neuroimagerie fonctionnelle : avancées conceptuelles et expérimentales

Jury :

  • ACHARD Sophie Directeur de Recherche, CNRS, GIPSA, Grenoble Rapporteur
  • HABRARD Amaury Professeur, Université de St Etienne, LHC Rapporteur
  • BONASTRE Jean-François Professeur, Université d’Avignon, LIA Examinateur
  • KADRI Hachem Associate Professeur, Aix-Marseille Université, LIS Examinateur
  • ARTIÈRES Thierry Professeur, LIS, Ecole Centrale de Marseille Directeur de thèse
  • TAKERKART Sylvain Ingénieur de Recherche, CNRS, INT Co-directeur de thèse

Abstract : In functional neuroimaging experiments, participants perform a set of tasks while their brain activity is recorded, e.g. with electroencephalography (EEG), magnetoencephalography (MEG) or functional magnetic resonance imaging (fMRI). Analysing data from a group of participants, which is often denoted as group-level analysis, aims at identifying traits in the data that relate with the tasks performed by the participant and that are invariant within the population. This allows understanding the functional organization of the brain in healthy subjects and its dysfunctions in pathological populations. While group-level analyses for classical univariate statistical inference schemes, such as the general linear model, have been heavily studied, there are still many open questions for group-level strategies based on multivariate machine learning methods. This thesis therefore focuses on multivariate group-level analysis of functional neuroimaging. We proposed a classifier-based multivariate group-level framework which we denote as inter-subject pattern analysis, We first performed a comparison of the results provided by inter-subject pattern analysis and the standard multivariate group-level strategy. Inter-subject analysis both offers a greater ability to detect statistically significant regions and facilitates the interpretation of the obtained results at a comparable computational cost. In this context, our second contribution is introducing an unifying formalization of inter-subject pattern analysis, that we cast as a multi-source transductive transfer learning problem, and then providing a survey of the methods where inter-subject pattern analysis was used in task-based functional neuroimaging experiments. Then we performed an experimental study that examines the well-foundedness of our multi-source transductive transfer formalization of intersubject pattern analysis. The fourth contribution of this thesis is a new multivariate group-level analysis method for functional neuroimaging datasets. Our method is based on optimal transport, which leverages the geometrical properties of multivariate brain patterns to overcome inter-individual differences impacting the traditional group-level analyses.

CNRS logo université Aix Marseille logo | plan du site | mentions légales | contact | admin | intranet | intcloud |