Multi-channel, Multi-slice, and Multi-contrast Compressed Sensing MRI UsingWeighted Forest Sparsity and Joint TV Regularization Priors

In Compressed Sensing based Magnetic Resonance Imaging (CS-MRI) reconstruction, wavelet transform finds the widest application as a sparsifying transform. It has been reported that along with the standard wavelet sparsity, group sparsity terms, like, the tree sparsity, the joint sparsity and the forest sparsity exist in the wavelet decomposition ofmulti-channel, multi-slice, and multi-contrastMRimages. To develop an efficient unified CS-MRI reconstruction model for these images, different wavelet based sparsity priors together plays a pivotal role. In this paper, we propose a simultaneous weighted forest sparsity and joint total variation based efficient CS-MRI reconstruction model for multi-channel, multi-slice, and multi-contrast MR images. The proposed technique shows significant improvements in terms of quality of reconstruction over existing methods for different datasets.