摘要
多线圈采集技术与并行成像算法通过降低磁共振成像所必需的梯度编码步数有效提高了成像的扫描速度。但是在数据采集过程中,运动常常会使线圈数据发生异常,从而对最终重建图像质量产生很大影响。本文提出了一种新的重建算法去消除重建图像中产生的伪影。算法把破坏数据看成观测数据样本中的异常值,应用了AM鲁棒估计进行数据修正,很好的抑制了异常值对数据集造成的影响。本研究分别对多线圈并行采集的体模数据与真实脑部数据进行了实验,结果显示算法可以有效消除破坏数据在重建图像中产生的伪影,并通过对比实验充分显示了本算法的优越性。
The speed of MR acquisition can be increased by decreasing the number of sequential phase encodes by means of arrays of multiple receiver coils. However, coil data are frequently corrupted due to motion during the data acquisition, which in turn have a significant on reconstructed composite image. In this article, a new algorithm is proposed to remove artifact in images. In this algorithm, corrupted data are regarded as outliers in observed datum. We apply robust Annealing M-Estimator to make solutions insensitive to the influence caused by outliers. Experimental results show the artifact cancellation for both phantom and in vivo. The benefits are demonstrated compared with SENSE techniques.
出处
《电路与系统学报》
CSCD
北大核心
2008年第1期18-22,共5页
Journal of Circuits and Systems
基金
国家"973"重点基础研究发展规划项目(2003CB716102)
