Wave-LORAKS: Combining wave encoding with structured low-rank matrix modeling for more highly accelerated 3D imaging.

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Image Reconstruction
• Accelerated Imaging

Background:

• Wave-CAIPI is an advanced technique for accelerating 3D MRI acquisition.
• LORAKS (Low-Rank Modeling of Localized Chemical Shifts) is a reconstruction framework utilizing data's low-rank properties.
• Previous LORAKS methods were limited to 2D reconstructions.

Purpose of the Study:

• To investigate the combination of Wave-CAIPI with LORAKS-based reconstruction (Wave-LORAKS).
• To enable further acceleration in 3D MRI beyond current Wave-CAIPI capabilities.
• To improve reconstruction quality and computational efficiency for 3D imaging.

Main Methods:

• Developed a 3D Wave-LORAKS reconstruction framework by integrating recent advances in structured low-rank matrix recovery.
• Retrospectively subsampled two fully sampled Wave-encoded 3D MPRAGE datasets.
• Compared Wave-LORAKS performance against traditional Wave-CAIPI reconstruction methods.

Main Results:

• Wave-LORAKS demonstrated superior reconstruction quality compared to Wave-CAIPI.
• Wave-LORAKS achieved higher acceleration factors, yielding better results with 16x accelerated data than Wave-CAIPI with 9x acceleration.
• The new method showed improved computational efficiency for large-scale 3D reconstructions.

Conclusions:

• Significant synergies exist between Wave encoding and LORAKS.
• Wave-LORAKS enables higher acceleration factors and more flexible sampling strategies in 3D MRI.
• This approach offers a promising advancement for accelerated 3D imaging acquisition and reconstruction.