An improved GRAPPA image reconstruction algorithm for parallel MRI

Reconstruction from partial k-space data is an important issue in parallel magnetic resonance imaging (PMRI), as k-space undersampling during data acquisition is liable to produce artifacts in the image. In order to remove the image aliasing due to k-space undersampling, this paper presents a new finite impulse response (FIR) model of GRAPPA algorithm to replace the FIR model whose coefficients are fixed and currently used in GRAPPA image reconstruction methods. The proposed FIR model has been a better description for the correlation of k-space data and a better approximation for the inversion of parallel imaging process. The method is demonstrated using the proposed GRAPPA algorithm with in vivo free-breathing cardiac imaging data and the results show that this improved algorithm can greatly improve the image quality even at very high acceleration factor.