题    目（TITLE）：Spatial regularization and sparsity for brain mapping

讲 座 人（SPEAKER）: Prof. Bertrand Thirion；INRIA

主 持 人 (CHAIR)： Prof. Shan Yu

时    间 (TIME)： 10:00AM, May19 (Thursday)

地    点 (VENUE)： The Second Meeting Room, 13th Floor

报告摘要（ABSTRACT）：

Inverse inference, or “brain reading”, is a recent paradigm for analyzing functional magnetic resonance imaging (fMRI) data, based on pattern recognition tools. By predicting some cognitive variables related to brain activation maps, this approach aims at decoding brain activity. Inverse inference takes into account the multivariate information between voxels and is currently the only way to assess how precisely some cognitive information is encoded by the activity of neural populations within the whole brain. However, it relies on a prediction function that is plagued by the curse of dimensionality, as we have far more features than samples, i.e., more voxels than fMRI volumes. To address this problem, different methods have been proposed. Among them are univariate feature selection, feature agglomeration and regularization techniques. We will give an overview of recently developed techniques to impose sparsity or compactness priors or predictive maps that seem particularly well suited to neuroimaging. In particular, when a generalization across individuals is of interest, more robustness to cross-individuals spatial variability may be achieved with adapted regularization or agglomeration methods. We will focus on the well-posedness of the estimation procedures and related optimization problems, then we will present tests of our algorithms on real datasets, and we show that the proposed algorithm yields better prediction accuracy than reference methods.

报告人简介（BIOGRAPHY）：

Bertrand Thirion is the principal investigator of the Parietal team (INRIA Saclay-Île-de-France) situated within the Neurospin research center at Saclay, France. After graduating from Ecole Polytechnique and Ecole Nationale Supérieure des Télécommunications, he got specialized in applied mathematics, with applications to computer vision. He did his PhD on the statistical analysis of functional brain images. His post doc at SHFJ, Orsay made him an active contributor of image analysis methods for between-subject comparison; since then he has been involved in neuroimaging platforms. Bertrand Thirion has been part of INRIA Saclay since 2006. His main research interests are the modeling of between brain variability in group studies, the mathematical study of functional connectivity and the use of machine learning tools for brain activity analysis; he addresses various applications such as the study of vision through neuroimaging, the classification of brain images for diagnosis or brain mapping and the study of correlations between neuroimaging and genetics information. He has co-authored about 50 papers in neuroimaging, medical image analysis and machine learning journals or conferences. He is part of the scientific board of the French neuroscience institute.