Compensation of Abdominal Breathing Artifacts in Multi-Shot Acquisitions Using Parallel Imaging and Automatic Coil Weight Selection from Adjacent Slices

Introduction Breathing motion during multi-shot acquisitions causes artifacts which appear as ghosts in the reconstructed images. Various techniques for motion correction using parallel imaging have been proposed [1-5]. They reconstruct images from subdivisions of motion corrupted k-space data and average them subsequently for artifact reduction [2], replace motion corrupted k-space lines [3] or correct for motion using parametric motion models [4, 5]. For multi-shot acquisitions of the abdomen these methods are of limited use: averaging may cause substantial blurring, often too much data is affected by motion to reject and replace it, and it is difficult to find a parametric model for breathing motion. Here parallel imaging and automatic coil weight selection are used to reconstruct images with reduced artifacts from undersampled k-spaces acquired during one shot of a multi-shot acquisition. Together with techniques for flexible image combination (e.g. image matching), that method can be used for breathing motion correction in abdominal imaging.