Vincent Michel

PhD in Computer Science 

In this thesis, I studied a technique for studying functional MRI data, called inverse inference, based on statistical learning methods. This technique is relatively recent (Cox et al. 2003; Mitchell et al. 2004), and has been used successfully in numerous studies in cognitive science, e.g. for the study of vision (Kamitani et al. 2005; Thirion and al. 2006), language (Mitchell et al. 2008), objects recognition (Reddy et al. 2003) and numbers representation (Knops et al. 2009).

I focused on the development of statistical learning algorithms which take into account the specificities of fMRI (in particular the spatial structure of images) to improve the performance obtained by inverse inference (see Haynes et al. 2006 for a state of the art of statistical learning techniques for inference reverse). I also applied these methods for solving specific neuroscientific problematics.

Research interests

• fMRI data analysis.
• Statistical learning, feature selection, predictive model, regularization.
• Spatial information, clustering.
•  Bayesian methods.
• Participation in the development of Scikit-learn, a library of statistical learning in Python.
  

Thesis

Phd Thesis : "Understanding the visual cortex by using classification techniques"

Peer-reviews publications 

• [1] V. Michel, C. Damon, and B. Thirion. Mutual information-based feature selection enhances fmri brain activity classification. In 5th Proc. IEEE ISBI, pages 592-595, 2008.

• [2] V. Michel, E. Eger, C. Keribin and B. Thirion. Adaptive multi-class bayesian sparse regression - an application to brain activity classification. In MICCAI'09 Workshop on Analysis of Functional Medical Images, 2009.
• [3] V. Michel, E. Eger, C. Keribin and B. Thirion. Multi-Class Sparse Bayesian Regression for Neuroimaging data analysis. In International Workshop on Machine Learning in Medical Imaging (MLMI) In conjunction with MICCAI 2010.

• [4] V. Michel, E. Eger, C. Keribin, J.-B. Poline and B. Thirion. A supervised clustering approach for extracting predictive information from brain activation images. In IEEE Computer Society Workshop on Mathematical Methods in Biomedical Image Analysis (MMBIA10) - IEEE Conference on Computer Vision and Pattern Recognition. 2010.
• [5] V. Michel, A. Gramfort, G. Varoquaux and B. Thirion. Total Variation regularization enhances regression-based brain activity prediction. In 1st ICPR Workshop on Brain Decoding - Pattern recognition challenges in neuroimaging - 20th International Conference on Pattern Recognition. 2010.
• [6] R. Genuer, V. Michel, E. Eger, and Thirion B. Random forests based feature selection for decoding fmri data. In COMPSTAT 19th International Conference on Computational Statistics, 2009.
• [7] M. Lebreton, S. Jorge, V. Michel, B. Thirion and M. Pessiglione. An automatic valuation system in the human brain : evidence from functional neuroimaging. Neuron 64, 3, 2009.
• [8] E. Eger, V. Michel, B. Thirion, A. Amadon, S. Dehaene and A. Kleinschmidt. Deciphering Cortical Number Coding from Human Brain Activity Patterns. Current Biology. 2009, 19:1608.
• [9] A. Knops, B. Thirion, E.M. Hubbard, V. Michel, S. Dehaene. Recruitment of an area involved in eye movements during mental arithmetic. Science. 2009 Jun 19;324(5934):1583-5.

Selected talks

MMBIA'10 - A supervised clustering approach for extracting predictive information from brain activation images.

ICPR'10 - Total Variation regularization enhances regression-based brain activity prediction.

MLMI'10 - Multi-Class Sparse Bayesian Regression for Neuroimaging data analysis.

Education

ESPCI (Ecole Supérieure de Physique et de Chimie Industrielles de Paris) engineer.
Master 2 in Applied Mathematics, Computer Vision and Machine Learning,  at the Ecole Normale Supérieure de Cachan (Cachan).
Teaching Unit in Anatomy and Imagery of the central nervous system at the Faculty of Medicine Pitié-Salpétrière (Paris).

Contact

Web: http://parietal.saclay.inria.fr/Members/vincent-michel

Address: Neurospin, Bâtiment 145, point courier 156,
                CEA Saclay, 91191 Gif sur Yvette – FRANCE

Phone: +33 (0)1 69 08 80 85,
Fax: +33 (0)1 69 08 79 80