3D multi-slab diffusion-weighted readout-segmented EPI with real-time cardiac-reordered k-space acquisition

Purpose: The aim of this study was to develop, implement, and demonstrate a three‐dimensional (3D) extension of the readout‐segmented echo‐planar imaging (rs‐EPI) sequence for diffusion imaging. Theory and Methods: Potential k‐space acquisition schemes were assessed by simulating their associated spatial point spread functions. Motion‐induced phase artifacts were also simulated to test navigator corrections and a real‐time reordering of the k‐space acquisition relative to the cardiac cycle. The cardiac reordering strategy preferentially chooses readout segments closer to the center of 3D k‐space during diastole. Motion‐induced phase artifacts were quantified by calculating the voxel‐wise temporal variation in a set of repeated diffusion‐weighted acquisitions. Based on the results of these simulations, a 2D navigated multi‐slab rs‐EPI sequence with real‐time cardiac reordering was implemented. The multi‐slab implementation enables signal‐to‐noise ratio‐optimal repetition times of 1–2 s. Results: Cardiac reordering was validated in simulations and in vivo using the multi‐slab rs‐EPI sequence. In comparisons with standard k‐space acquisitions, cardiac reordering was shown to reduce the variability due to motion‐induced phase artifacts by 30–50%. High‐resolution diffusion tensor imaging data acquired with the cardiac‐reordered multi‐slab rs‐EPI sequence are presented. Conclusion: A 3D multi‐slab rs‐EPI sequence with cardiac reordering has been demonstrated in vivo and is shown to provide high‐quality 3D diffusion‐weighted data sets.