A new method for removing motion artifacts in parallel MRI reconstruction

摘要

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 involuntary movement during the data acquisition, which in turn have a significant influence on reconstructed composite image. In this article, a new method called Weighted-Sum-of-Squares (WSOS) is proposed to remove artifacts in the reconstructed images. The method firstly adopts homomorphic filtering and B-spline smoothing to obtain a group of uniform brightness phased array coil images, and then accurately estimated corrupt samples based on mutual information between the coil images. Finally, weights are redistributed utilizing estimative knowledge in the following combined process. In this way, the influence of corrupt samples is attenuated so that motion artifacts in reconstructed images could be effectively reduced. The abdomen and brain experiments show that the new method reduces the motion artifacts caused by the corrupt data in the reconstructed images and the benefits are demonstrated as well by comparing it to conventional method.