An aliasing artifacts reducing approach with random undersampling for spatiotemporally encoded single-shot MRI
An aliasing artifacts reducing approach with random undersampling for spatiotemporally encoded single-shot MRI
复制标题
一种用于时空编码单次 MRI 的随机欠采样减少混叠伪影的方法
DOI：
10.1016/j.jmr.2013.10.005
复制
发表时间：
2013-12-01
影响因子：
2.2
通讯作者：
Chen, Zhong
中科院分区：
文献类型：
--
作者：
Chen, Lin;Bao, Lijun;Chen, Zhong
与回波平面成像(EPI)相比，时空编码(SPEN)单次激发成像具有更好的抗场不均匀性，同时在超分辨重建后保持了相当的空间和时间分辨率。虽然已经提出了各种重建方法，但由于大量的欠采样，重建的SPEN图像通常含有混叠伪影。提出了一种基于随机采样、奇异值分解(SVD)和压缩感知(CS)的混合算法来减少混叠伪影，提高图像质量。通过对水模和活体大鼠脑的数值模拟和实验，验证了该混合方法的有效性。本文提供的混合方案在大范围欠采样情况下将有利于SPEN方法。(C)2013 Elsevier Inc.保留所有权利。
Compared to the echo planar imaging (EPI), spatiotemporally encoded (SPEN) single-shot MRI holds better immunity to the field inhomogeneity, while retaining comparable spatial and temporal resolutions after the super-resolved reconstruction. Though various reconstruction methods have been proposed, the reconstructed SPEN images usually contain aliasing artifacts because of vast undersampling. A hybrid scheme based on random sampling, singular value decomposition (SVD) and compressed sensing (CS) was introduced to reduce these aliasing artifacts and improve the image quality. The efficiency of this hybrid scheme was demonstrated by numerical simulations and experiments on water phantom and in vivo rat brain. The hybrid scheme provided herein would benefit the SPEN approach in vast undersampiing situation. (C) 2013 Elsevier Inc. All rights reserved.