Extracting Brain Regions from Rest fMRI with Total-Variation Constrained Dictionary Learning

Alexandre Abraham^1,2, Elvis Dohmatob^1,2, Bertrand Thirion^1,2, Dimitris Samaras^3,4, and Gael Varoquaux^1,2

¹Parietal Team, INRIA Saclay-Île-de-France, Saclay, France
alexandre.abraham@inria.fr

²CEA, DSV, I2BM, Neurospin bât 145, 91191, Gif-Sur-Yvette, France

³Stony Brook University, NY 11794, USA

⁴Ecole Centrale, 92290, Châtenay Malabry, France

Abstract. Spontaneous brain activity reveals mechanisms of brain function and dysfunction. Its population-level statistical analysis based on functional images often relies on the definition of brain regions that must summarize efficiently the covariance structure between the multiple brain networks. In this paper, we extend a network-discovery approach, namely dictionary learning, to readily extract brain regions. To do so, we introduce a new tool drawing from clustering and linear decomposition methods by carefully crafting a penalty. Our approach automatically extracts regions from rest fMRI that better explain the data and are more stable across subjects than reference decomposition or clustering methods.

Keywords: dictionary learning, clustering, resting state fMRI

LNCS 8150, p. 607 ff.

Full article in PDF | BibTeX

lncs@springer.com
© Springer-Verlag Berlin Heidelberg 2013