Motivation: Previously proposed iSLIPEN is a promising technique to avoid slab-boundary-artifacts in 3D multi-slab DWI. However, it suffers from manifest resolution loss and residual aliasing because of using 2D partial-Fourier and regular under-sampling pattern along ky direction. Goal(s): To propose a modified sampling-pattern and refine the reconstruction-model for iSLIPEN to recover the resolution and reduce the residual aliasing. Approach: Structured-low-rank and 2D-CAIPIRINHA were incorporated into iSLIPEN framework to reconstruct the data acquired with only one third of gold standard. Results: The modified iSLIPEN can not only avoid slab-boundary-artifacts, but can also recover the resolution and reduce residual aliasing, with comparable quality to gold standard. Impact: 3D isotropic high-resolution DWI without suffering from slab boundary artifacts, loss of resolution and residual aliasing can be robustly achieved by our method, therefore benefiting the neuroscience study in evaluating crossing and kissing fibers.
Xu et al. (Tue,) studied this question.