4.4 Article

Single-shell to multi-shell dMRI transformation using spatial and volumetric multilevel hierarchical reconstruction framework

Journal

MAGNETIC RESONANCE IMAGING
Volume 87, Issue -, Pages 133-156

Publisher

ELSEVIER SCIENCE INC
DOI: 10.1016/j.mri.2021.12.011

Keywords

MRI; Diffusion MRI; Single-shell HARDI; Multi-shell HARDI; Spherical Harmonics; fODF; GANs; Attention Module; Encoder-Decoder N; W

Funding

  1. McDonnell Center for Neuroscience Systems at the University of Washington
  2. [1U54MH091657]

Ask authors/readers for more resources

Single or multi-shell high angular resolution diffusion imaging is an important technique for studying brain white matter fibers. The existing single-shell technique has limitations in estimating the intravoxel structure at the desired resolution. In this study, a deep learning architecture is proposed to reconstruct diffusion MRI volumes for different b-values using single-shell acquisitions, allowing for higher resolution and accurate fiber tracts. The proposed framework incorporates contextual information within each slice and across the slices to optimize network learning, achieving promising results in the validation.
Single or Multi-shell high angular resolution diffusion imaging (HARDI) has become an important dMRI acquisition technique for studying brain white matter fibers. Existing single-shell HARDI makes it challenging to estimate the intravoxel structure up to the desired resolution. However, multi-shell acquisition (with multiple bvalues) can provide higher resolution for the intravoxel structure, which further helps in getting accurate fiber tracts; But, this comes at the cost of larger acquisition time and larger setup. Hence, we propose a novel deep learning architecture for the reconstruction of diffusion MRI volumes for different b-values (degree of diffusion weighting) using acquisitions at a fixed b-value (termed as single-shell) acquisition. This reconstruction has been performed in the spherical harmonics space to better manage varying gradient directions. In this work, we have demonstrated such a reconstruction for b = 3000 s/mm2 and b = 2000 s/mm2 from b = 1000 s/mm2. The proposed Multilevel Hierarchical Spherical Harmonics Coefficients Reconstruction (MHSH) framework takes advantage of contextual information within each slice as well as across the slices by involving Slice Level ReconNet (SLRNet) network and a Volumetric ROI Level ReconNet (VPLRNet) network, respectively. Three-loss functions have been used to optimize network learning, i.e., L1, Adversarial, and Total Variation Loss. Finally, the network is trained and validated on the publicly available HCP data-set with standard qualitative and quantitative performance measures and achieves promising results.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.4
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available