Technical - Emerging CMR Methods - Acquisition & Reconstruction-Slice low-rank plus SPARSE (slice-L+S) outperforms slice-SPARSE-SENSE and sequential operations for reconstruction of k-t undersampled multiband first-pass myocardial perfusion MRI

Slice low-rank plus SPARSE (slice-L+S) outperforms slice-SPARSE-SENSE and sequential operations for reconstruction of k-t undersampled multiband first-pass myocardial perfusion MRI

Back to course

Video Transcription

Video Summary

The presentation describes a new reconstruction method for KT-undersampled multiband first-pass myocardial perfusion MRI, which aims to improve heart coverage while reducing artifacts. The proposed Slice Low-Rank Plus Spars method combines data consistency with temporal low-rank/sparsity constraints in a single optimization framework. Compared with Slice-Spars SENSE and sequential reconstruction methods, it reduces slice leakage, undersampling artifacts, and blurring. Results from retrospective and prospective patient studies show better image quality, temporal fidelity, and quantitative performance. The authors conclude that their approach outperforms prior methods and may be extended with different acceleration factors and deep learning in future work.

Keywords

myocardial perfusion MRI

KT-undersampled multiband

Slice Low-Rank Plus Spars

image reconstruction

undersampling artifacts