%0 Journal Article
%T Accelerating Dynamic Cardiac MR Imaging Using Structured Sparse Representation
%A Nian Cai
%A Shengru Wang
%A Shasha Zhu
%A Dong Liang
%J Computational and Mathematical Methods in Medicine
%D 2013
%I Hindawi Publishing Corporation
%R 10.1155/2013/160139
%X Compressed sensing (CS) has produced promising results on dynamic cardiac MR imaging by exploiting the sparsity in image series. In this paper, we propose a new method to improve the CS reconstruction for dynamic cardiac MRI based on the theory of structured sparse representation. The proposed method user the PCA subdictionaries for adaptive sparse representation and suppresses the sparse coding noise to obtain good reconstructions. An accelerated iterative shrinkage algorithm is used to solve the optimization problem and achieve a fast convergence rate. Experimental results demonstrate that the proposed method improves the reconstruction quality of dynamic cardiac cine MRI over the state-of-the-art CS method. 1. Introduction Dynamic cardiac cine MR imaging aims at simultaneously providing a series of dynamic magnetic resonance image in spatial and temporal domains ( space) at a high frame rate. It usually acquires the -space at each time frame and collects the raw data in the spatial frequency and temporal domain, the so called space. Therefore, it is necessary to reconstruct each time frame and get a series of dynamic images. However, the relatively low acquisition speed of the dynamic MRI is an important factor to limit its application in clinics. How to accelerate -space sampling for each time frame and reconstruct them without sacrificing spatial resolution is a challenging problem. In recent years, many advanced techniques [1¨C10] were proposed to effectively address this issue and can be divided into two categories. One is based on compressed sensing (CS) theory [11, 12] utilizing the sparsity in dynamic images to be reconstructed, and the other is based on the partial separable theory [13] exploiting the low-rank property of images in space. The application of CS in dynamic MRI has drawn a lot of attention, since this theory demonstrates that the signal can be accurately reconstructed from a small amount of linear undersampled measurements by exploiting the inherent sparsity in signal. For example, Jung et al. [7, 9] uncovered an intriguing link between the compressed sensing and BLAST/SENSE and proposed the FOCUSS algorithm to achieve high spatiotemporal resolution in cardiac cine imaging. Liang et al. [5] developed iterative support detection ( ISD) method to further utilize the detected partial support information besides the sparsity in cardiac cine images. Recently, image restoration with patch-based sparse representations has attracted a lot of attention. The similarity of works in this topic is seeking for a more appropriate way to
%U http://www.hindawi.com/journals/cmmm/2013/160139/