Sparsely sampled functional magnetic resonance imaging using low-rank and sparsity constraints

Sparse sampling has been demonstrated useful for spatiotemporal imaging. Functional MRI benefits from sparse sampling due to reduction of the readout duration which in turn makes the acquired data less prone to T₂* susceptibility artifacts caused by magnetic field gradients near air-tissue interfaces. Conventional Fourier transform-based reconstruction method suffers from sparse sampling artifacts such as high-frequency ringing and loss of resolution. To address the problem of sparse sampling and field inhomogeneity artifacts at once, we propose to exploit the specific low-rank structure of fMRI data via group sparsity and incorporate field inhomogeneity into iterative image reconstruction. Experimental results demonstrate the ability of the method in obtaining higher-resolution functional images and activation maps with artifact correction due to susceptibility-induced field gradients.