Showing papers by "Bertrand Thirion published in 2003"

Feature Detection in fMRI Data: The Information Bottleneck Approach

Abstract: Clustering is a well-known technique for the analysis of fMRI data, whose main advantage is certainly flexibility: given a metric on the dataset, it defines the main features contained in the data But intrinsic to this approach are also the problem of defining correctly the quantization accuracy, and the number of clusters necessary to describe the data The Information Bottleneck (IB) approach to vector quantization [11] addresses these difficulties: 1) it deals with an explicit tradeoff between quantization and data fidelity; 2) it does so during the clustering procedure and not post hoc; 3) it takes into account the statistical distribution of the features within the feature space and not only their most likely value; last, it is principled through an information theoretic formulation, which is relevant in many situations In this paper, we present how to benefit from this method to analyze fMRI data Our application is the clustering of voxels according to the magnitude of their responses to several experimental conditions The IB quantization provides a consistent representation of the data, allowing for an easy interpretation

fMRI data analysis : statistics, information and dynamics

Abstract: In this thesis, we discuss and propose several methods for functional MRI -magnetic resonance imaging- data analysis. Functional MRI is a recent modality for the study of brain function: it produces image sequences that reflect local brain metabolic activity, which in turn reflects neural activity. We first deal with the modeling of each voxel-based temporal pattern, using linear prediction techniques and estimating the information contained in the temporal processes. We then study different multivariate generalizations of this model. After recalling and discussing several classical methods (independent components analysis, clustering), we first propose a linear approach based on state-space modeling, and then a non-linear approach based on kernel decompositions. The common objective of these methods -that are nevertheless complementary- is to propose decompositions that preserve optimally the data dynamics. We then introduce a novel point of view based on dimension reduction of the dataset, which allows for a more structured representation and helps for visualization. We show its effectiveness with respect to classical linear decomposition techniques. Finally, we describe a methodology of analysis that synthesizes different parts of this work, and relies on soft hypotheses. Our results give a global description of the dynamical processes embedded in the data produced by functional MRI.

Analyse de données d' IRM fonctionnelle : statistiques, information et dynamique

Abstract: Dans cette these, nous discutons et proposons un certains nombre de methodes pour l'analyse de donnees d'IRM -imagerie par resonance magnetique- fonctionnelle. L'IRM fonctionnelle est une modalite recente de l'exploration du cerveau: elle produit des sequences d'images refletant l'activite metabolique locale, celle-ci refletant l'activite neuronale. Nous nous interessons tout d'abord a la modelisation des series temporelles obtenues pour chaque voxel separement, en faisant appel aux techniques de prediction lineaire et au calcul de l'information des processus modelises. Nous etudions ensuite differentes generalisations multivariees de ce modele. Apres avoir rappele et discute certaines techniques classiques (analyse en composantes independantes, regroupement), nous proposons successivement une approche lineaire fondee sur la theorie des systemes a etat et une approche non-lineaire fondee sur les decompositions a noyau. Le but commun de ces methodes -qui peuvent se completer- est de proposer des decompositions qui preservent au mieux la dynamique des donnees. Nous introduisons ensuite une approche nouvelle par reduction de la dimension des donnees; cette approche offre une representation plus structuree et relativement agreable a visualiser. Nous montrons ses avantages par rapport aux techniques lineaires classiques. Enfin, nous decrivons une methodologie d'analyse qui synthetise une grande partie de ce travail, et repose sur des hypotheses tres souples. Nos resultats offrent ainsi une description globale des processus dynamiques qui sont mis en image lors des experiences d'IRM fonctionnelle

Dynamical Components Analysis of FMRI Data: A Second Order Solution

Abstract: In this paper, we present a new way to derive low dimensional representations of functional MRI datasets. This is done by introducing a state-space formalism, where the state corresponds to the components whose dynamical structure is of interest. The rank of the selected state space is chosen by statistical comparison with null datasets. We study the validity of our estimation scheme on a synthetic dataset, and show on a real dataset how the interpretation of the complex dynamics of fMRI data is facilitated by the use of low-dimensional, denoised representations. This methods makes a minimal use of priors on the data structure, so that it is very practical for exploratory data analysis.