An efficient interpolated compressed sensing method for highly correlated 2D multi-slice MRI

In 2D multi-slice magnetic resonance imaging (MRI), adjacent slices are normally highly correlated which can be exploited for non-uniform random undersampling. To compensate for undersampling, missing samples of intermediate slices may be estimated from their neighboring slices. In this paper, we propose an efficient two phase 2D multi-slice MR image reconstruction technique combining interpolation with wavelet tree sparsity based CS reconstruction from highly undersampled measurements. Simulation results show that the proposed method gives average improvements of 1.8-3.6 dB in reconstructions compared to the state-of-the-art within a clinically significant time.