《3D WAVELET-BASED REGULARIZATION FOR PARALLEL MRI》.pdf

3D WAVELET-BASED REGULARIZATION FOR PARALLEL MRI RECONSTRUCTION: IMPACT ON SUBJECT AND GROUP-LEVEL STATISTICAL SENSITIVITY IN FMRI 1,2 2 2 2 1 Lotfi Chaari , S´ebastien M´eriaux , Solveig Badillo , Philippe Ciuciu and Jean-Christophe Pesquet 1 2 2 Universit´e Paris-Est, IGM and UMR-CNRS 8049, CEA/DSV/I BM/Neurospin, CEA Saclay, Bbt. 145, 77454 Marne-la-Vall´ee cedex, France Point Courrier 156, 91191 Gif-sur-Yvette cedex, France {chaari,pesquet}@univ-mlv.fr {sebastien.meriaux,solveig.badillo,philippe.ciuciu}@cea.fr ABSTRACT have been proposed in the literature to improve the robust- Parallel MRI is a fast imaging technique that allows recon- ness of the solution [1, 4–6]. Some of them apply quadratic struction of full Field-of-View (FoV) images based on under- regularization while others resort to edge-preserving regu- sampled k-space data acquired using multiple receiver coils larization in the wavelet transform domain. To the best of with complementary sensitivity profiles. It enables the ac- our knowledge only one study has been conducted to assess quisition of highly resolved images either in space or in time, the impact of using a specific pMRI reconstruction algorithm which is of particular interest in applications like functional on brain activation detection in fMRI [7]. There are several neuroimaging. Thes