一种基于压缩感知的MR成像的新方法

A New Method of MR Imaging Based on Compressed Sensing

FCSA(Fast Composite Splitting Algorithm)是一种用压缩感知理论进行MR重建的有效算法,它在重构时间和图像重构质量上相比其他算法有明显优势.但是,当算法选择不同的参数时,重建效果不稳定.本文对该算法进行了改进,以减少由人为选择不同的参数所带来的对重建效果的影响.实验表明,改进后的算法在参数选择不同时,重建效果都能够得到保证.

FCSA （Fast Composite Splitting Algorithm） is an efficient algorithm to reconstruct MR images based on compressed sensing theory. It has obvious advantages in reconstruction time and reconstructed image quality when compared with some other algorithms. However, when different parameters are chosen in this algorithm, the reconstructed image quality is inconsistent. In this paper, some improvement of FCSA is made in order to reduce the negative effect of reconstruc- ted image quality resulted from different personal choices of parameters. Experiments show that the reconstructed image quality can be guaranteed by using the improved algorithm.