Autoregressive Moving Average (ARMA) Method for the Reconstruction of MR Images from Sparsely Sampled 3D k-Space
Peng H, Smith M, Frayne R
University of Calgary, Seaman Family MR Research Centre, Foothills Medical Centre, Calgary Health Region
In order to achieve real-time 3D MR imaging one approach is to collect sparsely sampled k-space datasets. Moving-table contrast-enhanced MR angiography is one such example. Here, we compare different reconstruction of sparsely sampled data by zero filling (ZF) and by TERA. The quality of the reconstructed images was assessed by visual inspection, as well as by both global and local performance errors (PE). It is shown that the proposed CP-TERA results in better image quality than ZF and rTERA from sparsely sampled k-space dataset and that the PE (both global and local) values in CP-TERA image were less than in ZF and rTERA images.