Abstract
Undersampled magnetic resonance image reconstruction employing sparsity regularization has fascinated many researchers in recent years under the support of compressed sensing theory. Nevertheless, most existing sparsity-regularized reconstruction methods either lack adaptability to capture the structure information or suffer from high computational load. With the aim of further improving image reconstruction accuracy without introducing too much computation, this paper proposes a data-driven tight frame magnetic image reconstruction (DDTF-MRI) method. By taking advantage of the efficiency and effectiveness of data-driven tight frame, DDTF-MRI trains an adaptive tight frame to sparsify the to-be-reconstructed MR image. Furthermore, a two-level Bregman iteration algorithm has been developed to solve the proposed model. The proposed method has been compared to two state-of-the-art methods on four datasets and encouraging performances have been achieved by DDTF-MRI.
Highlights
Magnetic Resonance Imaging (MRI) is one of the major diagnostic imaging modalities with noninvasive and nonionizing radiation nature
We evaluated the proposed DDTFMRI on four datasets and compared it to two state-of-theart methods, including CSMRI-total variation (TV) and DLMRI [5]
A data-driven tight frame (DDTF)-MRI method has been proposed in this paper to effectively reconstruct MR image from undersampled K-space data
Summary
Magnetic Resonance Imaging (MRI) is one of the major diagnostic imaging modalities with noninvasive and nonionizing radiation nature. The CS theory has shown that if an image has a sparse representation under certain transform, we can precisely restore the original image from its partial measurements under the RIP condition [3, 4] With such a transform, the MR image reconstruction from its undersampled Kspace data can be achieved by nonlinear algorithms, like l1 minimization or orthogonal match pursuit (OMP) algorithm [2, 5, 6]. The proposed method relies on a perfect reference image, which is quite hard to be obtained from its undersampled Kspace data, and the data-driven tight frame is learnt from the reference image instead of the target image.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
