Identifying predictive regions from fMRI with TV-L1 prior

Abstract : Decoding, i.e. predicting stimulus related quantities from functional brain images, is a powerful tool to demonstrate differences between brain activity across conditions. However, unlike standard brain mapping, it offers no guaranties on the localization of this information. Here, we consider decoding as a statistical estimation problem and show that injecting a spatial segmentation prior leads to unmatched performance in recovering predictive regions. Specifically, we use L1 penalization to set voxels to zero and Total-Variation (TV) penalization to segment regions. Our contribution is two-fold. On the one hand, we show via extensive experiments that, amongst a large selection of decoding and brain-mapping strategies, TV+L1 leads to best region recovery. On the other hand, we consider implementation issues related to this estimator. To tackle efficiently this joint prediction-segmentation problem we introduce a fast optimization algorithm based on a primal-dual approach. We also tackle automatic setting of hyper-parameters and fast computation of image operation on the irregular masks that arise in brain imaging.
Document type :
Conference papers
Complete list of metadatas

Cited literature [11 references]  Display  Hide  Download

https://hal.inria.fr/hal-00839984
Contributor : Alexandre Gramfort <>
Submitted on : Monday, July 1, 2013 - 12:11:59 PM
Last modification on : Thursday, March 7, 2019 - 3:34:14 PM
Long-term archiving on : Wednesday, October 2, 2013 - 4:12:23 AM

Files

paper.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00839984, version 1

Citation

Alexandre Gramfort, Bertrand Thirion, Gaël Varoquaux. Identifying predictive regions from fMRI with TV-L1 prior. Pattern Recognition in Neuroimaging (PRNI), Jun 2013, Philadelphia, United States. ⟨hal-00839984⟩

Share

Metrics

Record views

2656

Files downloads

878