一种新的基于GRAPPA的图像重建算法

A New Algorithm for Image Reconstruction Based on GRAPPA

传统GRAPPA(generalized auto-calibrating partially parallel acquisitions)采样轨迹存在重建图像质量差、数据扫描时间长等问题。在传统GRAPPA基础上提出了一种新的图像重建算法——交叉采样轨迹法,利用在相位编码方向上线圈间的空间灵敏度信息的互补特性,能够得到准确的线圈权重系数,提高了重建图像的质量。通过对真实脑部数据进行实验,结果显示该方法相比于传统的GRAPPA采样轨迹,可提高重构图像的质量、缩短数据扫描时间,是一种有效的采样方法。

The traditional sampling trajectory of GRAPPA（generalized auto-calibrating partially parallel acquisitions）has some problems, such as bad reconstruction image, long data scanning time and so on. A new algorithm for image reconstruction based on GRAPPA was proposed, which called crossed sampling trajectory. This method takes advantage of the complementary character between coils in spatial sensitivity information in the phase encoding direction, gets precise coil weight coefficients, and improves the recon- struction image＇s quality. Brain imaging experiment using real brain data demonstrated that, compared with traditional GRAPPA trajectory, this new method not only improved quality of reconstruction image but also reduced time of data scanning. In a word, crossed sampling trajectory is an effective method.