Cross-sampled GRAPPA for parallel MRI.

Area of Science:

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

Background:

• Generalized Auto-calibrating Partially Parallel Acquisitions (GRAPPA) is a key parallel imaging technique.
• GRAPPA reconstructs undersampled k-space data using acquired data and calibration weights.
• Current GRAPPA methods derive weights from auto-calibration signal (ACS) lines parallel to undersampled data.

Purpose of the Study:

• To introduce a novel cross-sampling method for acquiring ACS data in GRAPPA.
• To enhance calibration accuracy and reduce aliasing artifacts in parallel MRI.
• To improve GRAPPA performance, particularly under high acceleration conditions.

Main Methods:

• Proposed a cross-sampling strategy to acquire ACS lines orthogonal to reduced k-space lines.
• Implemented the cross-sampled GRAPPA (CS-GRAPPA) technique.
• Validated the method using both phantom and in vivo MRI experiments.

Main Results:

• The cross-sampling method increases calibration data in the undersampled direction.
• CS-GRAPPA demonstrated improved calibration accuracy compared to conventional GRAPPA.
• Significant reduction in aliasing artifacts was observed with CS-GRAPPA, especially at high acceleration.

Conclusions:

• Cross-sampled GRAPPA (CS-GRAPPA) offers enhanced performance for parallel MRI.
• The proposed ACS acquisition strategy is effective in improving GRAPPA's robustness.
• CS-GRAPPA is a valuable advancement for achieving higher acceleration factors in MRI.