Improved multi-shot diffusion imaging using GRAPPA with a compact kernel.

Area of Science:

• Magnetic Resonance Imaging
• Image Reconstruction
• Diffusion Imaging

Background:

• Motion-induced phase variations in multi-shot diffusion imaging cause artifacts.
• Traditional SENSE-based corrections require geometric matching of image and navigator echoes.
• Existing k-space methods like realigned GRAPPA offer an alternative approach.

Purpose of the Study:

• To refine the theory and application of GRAPPA-based methods for motion artifact correction in diffusion imaging.
• To develop a more robust reconstruction technique less sensitive to geometric mismatches.

Main Methods:

• Revised GRAPPA theory treating phase variations as encoding, analogous to coil sensitivity encoding.
• Synthesis of missing data using k-space correlations across shots and channels.
• Utilized a compact kernel focusing on significant acquired data for accurate weight generation.

Main Results:

• The proposed method effectively suppresses artifacts from phase variations.
• Achieved high-resolution diffusion images with reduced distortion.
• Demonstrated robustness in simulation and experimental (brain, cervical spine) data.

Conclusions:

• The refined GRAPPA-based method offers superior artifact suppression and image quality in diffusion imaging.
• This approach is less sensitive to geometric mismatches between image and navigator echoes compared to SENSE.
• Provides a promising alternative for motion artifact correction in challenging diffusion MRI acquisitions.