Motion-corrected independent component analysis (MCICA) of functional magnetic resonance imaging time series

A 3D image registration method for motion correction of functional magnetic resonance imaging (fMRI) time-series, based on Independent Component Analysis (ICA), is described. We argue that movement during fMRI data acquisition results in a simultaneous increase in the joint entropy of the observed time-series and a decrease in the joint entropy of a nonlinear function of the derived spatially independent components calculated by ICA. We propose this entropy difference as a reliable criterion for motion correction and refer to a method that maximizes this as Motion-Corrected ICA (MCICA). Specifically, a given motion-corrupted volume may be corrected by determining the linear combination of spatial transformations of the motion-corrupted volume that maximizes the proposed criterion. In essence, MCICA consists of designing an adaptive spatial resampling filter which maintains maximum temporal independence among the recovered components. Simulations demonstrate that MCICA is robust to activation level, additive noise, random motion in the reference volumes and the exact number of independent components extracted. When MCICA is compared with the conventional square of difference-based measures like LS-SPM and LS-AIR, we demonstrate that in simulations, MCICA is more robust to the addition of simulated activation and dose not lead to detection of false activations after correction of simulated task-correlated motion. With actual data from a motor fMRI experiment, the time course of the derived continually task-related ICA component become significantly more correlated with the underlying behavioral task after preprocessing with MCICA compared to other methods, and the associated activation map is more clustered in the primary motor and supplementary motor cortices without spurious activation at the brain edge. When the variance of the representative motion-corrected data is compared to original data, 80% of voxels decreased their variance with MCICA, compared with 64% (SPM) and 67% (AIR). Although Mutual Information is not explicitly optimized, the MI between all subsequent volumes and the first one is consistently increased after preprocessing the data with MCICA. We suggest that by assessing the statistical properties of a motion-corrupted volume in relation to other volumes in the series, MCICA provides a robust and reliable method for preprocessing of fMRI time-series corrupted with motion.