Calibrationless Parallel MRI Reconstruction by Using Joint Sparsity Feature
Abstract— In parallel magnetic resonance imaging (pMRI), images between different coils have the similar location of the singularities or discontinuities. As sparsifying transform captures the discontinuities in the images and can be assumed not to affect the position of the discontinuities in the coil images, the corresponding transform results of images from multi-coils can be considered as joint sparse. But previous methods do not consider the property for MRI reconstruction when they include both wavelet transform and total variation. In this paper, we propose a new method based on fast iterative shrinkage/thresholding algorithm (FISTA) and split Bregman algorithm to reconstruct multi-coils images, in which it contains both the joint wavelet sparsity (JWS) and the joint total variation (JTV) regularizers. The experimental results of phantom and brain images show that our proposed algorithm performs better than the other state-of-the-art algorithms.
Index Terms— compressed sensing (CS), pMRI, JWS, JTV, FISTA, split bregman.
Jiaquan Jin, Hongwei Du, Bensheng Qiu
Center for Biomedical Imaging of University of Science and Technology of China, CHINA
School of Electrical Engineering and Automation, Hefei University of Technology, CHINA
Cite: Jiaquan Jin, Hongwei Du, Bensheng Qiu, Jinzhang Xu, "Calibrationless Parallel MRI Reconstruction by Using Joint Sparsity Feature," Proceedings of 2018 the 8th International Workshop on Computer Science and Engineering, pp. 182-186, Bangkok, 28-30 June, 2018.