‐corrected water–fat imaging using compressed sensing and parallel imaging

Abstract

To demonstrate an approach to water-fat separation with R2* correction using compressed sensing and parallel imaging. Acquisition times for chemical shift based water-fat separation imaging are lengthy, and many applications rely on image acceleration techniques. In this study, we present an integrated compressed sensing, parallel imaging, R2* corrected water-fat separation technique for water-fat imaging of highly accelerated acquisitions. Reconstruction times are reduced using coil compression. The proposed technique is demonstrated using a customized IDEAL-SPGR pulse sequence to acquire retrospectively and prospectively undersampled datasets of the liver, calf, knee, and abdominal cavity. This technique is shown to offer comparable image quality relative to fully sampled reference images for a range of acceleration factors. At high acceleration factors, this technique is shown to offer improved image quality over parallel imaging. A technique is described that uses compressed sensing and parallel imaging to reconstruct R2*-corrected water and fat images from accelerated datasets. Acceleration factors as high as 7.0 are shown with excellent image quality. These high acceleration factors enable water-fat separation with higher resolution or greater anatomical coverage in breath-hold applications.