Meeting Banner
Abstract #3274

k-T FASTER: A New Method for the Acceleration of Resting State fMRI Data Acquisition

Mark Chiew1, Stephen M. Smith1, Peter J. Koopmans2, Thomas Blumensath3, Karla L. Miller1

1FMRIB Centre, University of Oxford, Oxford, United Kingdom; 2Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen, Nijmegen, Netherlands; 3ISVR, University of Southampton, Southampton, Hampshire, United Kingdom

In FMRI, measurements of resting state functional connectivity are often preceded by a principal component analysis to reduce data dimensionality. We propose a new method for the acceleration of FMRI acquisitions that exploits the decrease of information in a dimensionality reduction to facilitate the undersampling of k-t space. We call this approach k-t FASTER: FMRI Acceleration in Space-time via Truncation of Effective Rank. This technique is demonstrated on 4x retrospectively undersampled FMRI data to reproduce resting state networks with high spatial fidelity.

Abstract PDF Presentation Video

Keywords

accelerated accelerating acceleration acquisition agreement always appropriate approximation approximations background blue bottom cardiac coefficients cognition coil collected combines comp companion completion component computed constitute continuation corrected correlation cortex dataset datasets decimated degrees dimension dimensionality dimensions domain domains dual dynamic efficiency entirely equivalent estimating every example excellent explore fast faster feasibility fidelity fixed freedom full functional fundamentally funded future grant ground hard harvest highest highlighting identified identify improved incorporation independent indicated indicating iterative kindly kingdom largely limited maps mark math matrices matrix miller minimization mixture network networks novel nuclear null observation original oxford parietal particularly performance peter poor positive previously principal prior produce produced propose proposed prospectively pulse randomly rank ranks recent reconstruct reconstructed reconstruction recovers recovery reduction regression related remains representative represented represents resolution respectively resting resurgence retrospectively sampled sampling score scores selecting sensing simulated smith smoothness space sparsity spatial suggest supplied table take temporal thresholding trans transform truncation truth yellow