A Slice-Low-Rank Plus Sparse (slice-L + S) Reconstruction Method for k-t Undersampled Multiband First-Pass Myocardial...

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Imaging Physics

Background:

• Multiband (MB) imaging combined with k-t undersampling can improve slice coverage and spatial resolution in first-pass perfusion MRI.
• The low-rank plus sparse (L+S) model is effective for accelerating single-band (SB) perfusion MRI.

Purpose of the Study:

• To develop and evaluate a novel data consistency method for MB perfusion MRI.
• To combine MB data consistency with temporal L+S constraints for improved reconstruction.

Main Methods:

• A MB data consistency method using ESPIRiT maps and through-plane coil information was developed.
• This method, termed slice-L+S, was integrated with temporal L+S constraints.
• Slice-L+S was compared against SB L+S and sequential split slice-GRAPPA with SB L+S (seq-SG-L+S) using synthetic and prospectively acquired MB data.

Main Results:

• Slice-L+S demonstrated superior performance over SB L+S and seq-SG-L+S in terms of normalized RMSE and structural similarity index on synthetic data (P<0.05).
• Reconstructions using slice-L+S showed better image quality compared to seq-SG-L+S for prospectively undersampled MB data, as assessed by blinded experts.
• Specific quantitative results showed slice-L+S achieving lower RMSE and higher structural similarity index values.

Conclusions:

• The developed slice-L+S method effectively reconstructs accelerated MB perfusion MRI data.
• Slice-L+S provides significant improvements in image quality and slice coverage for cardiac MRI.
• This method offers a promising approach for advanced perfusion imaging with high spatial resolution.