Total Variation regularization enhances regression-based brain activity prediction

Abstract : While medical imaging typically provides massive amounts of data, the automatic extraction of relevant information in a given applicative context remains a difficult challenge in general. With functional MRI (fMRI), the data provide an indirect measurement of brain activity, that can be related to behavioral information. It is now standard to formulate this relation as a machine learning problem where the signal from the entire brain is used to predict a target, typically a behavioral variable. In order to cope with the high dimensionality of the data, the learning method requires a regularization procedure. Among other alternatives, L1 regularization achieves simultaneously a selection of the most predictive features. One limitation of the latter method, also referred to as Lasso in the case of regression, is that the spatial structure of the image is not taken into account, so that the extracted features are often hard to interpret. To obtain more informative and interpretable results, we propose to use the `1 norm of the image gradient, a.k.a., the Total Variation (TV), as regularization. TV extracts few predictive regions with piecewise constant weights over the whole brain, and is thus more consistent with traditional brain mapping. We show on real fMRI data that this method yields more accurate predictions in inter-subject analysis compared to voxel-based reference methods, such as Elastic net or Support Vector Regression.
Type de document :
Communication dans un congrès
1st ICPR Workshop on Brain Decoding - Pattern recognition challenges in neuroimaging - 20th International Conference on Pattern Recognition, Aug 2010, Istanbul, Turkey. IEEE Computer Society, 0, pp.9-12, 2010, 〈10.1109/WBD.2010.13〉
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-00504095
Contributeur : Vincent Michel <>
Soumis le : lundi 19 juillet 2010 - 21:01:22
Dernière modification le : jeudi 7 février 2019 - 14:35:35
Document(s) archivé(s) le : mardi 23 octobre 2012 - 10:41:26

Fichier

icpr_2010_tv.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Vincent Michel, Alexandre Gramfort, Gaël Varoquaux, Bertrand Thirion. Total Variation regularization enhances regression-based brain activity prediction. 1st ICPR Workshop on Brain Decoding - Pattern recognition challenges in neuroimaging - 20th International Conference on Pattern Recognition, Aug 2010, Istanbul, Turkey. IEEE Computer Society, 0, pp.9-12, 2010, 〈10.1109/WBD.2010.13〉. 〈hal-00504095〉

Partager

Métriques

Consultations de la notice

578

Téléchargements de fichiers

272