Motion Correction in High-Resolution 3D Brain MRSI Without Water and Lipid Suppression

Area of Science:

• Magnetic Resonance Imaging
• Neuroimaging
• Biomedical Engineering

Background:

• Magnetic Resonance Spectroscopic Imaging (MRSI) is crucial for in vivo metabolite quantification.
• Long acquisition times in MRSI make it highly susceptible to subject motion.
• Non-water-suppressed MRSI presents unique challenges for motion correction due to artifacts in water and lipid signals.

Purpose of the Study:

• To develop an effective motion correction method for high-resolution, non-water-suppressed MRSI.
• To address the challenges posed by motion artifacts in both water/lipid signals and metabolite quantification.
• To enhance the robustness and clinical utility of MRSI techniques.

Main Methods:

• A novel motion correction approach using TR-wise linear navigators to detect and discard motion-corrupted k-space data.
• Reconstruction of discarded data by treating motion correction as a missing data problem, incorporating sensitivity encoding.
• Integration of extrinsic (water/lipid) and intrinsic (metabolite) signal priors into the correction process.
• Validation using a spectroscopic phantom and human subjects, including healthy adults, a child, and a patient with involuntary movements.

Main Results:

• The method achieved high-quality water images and metabolite maps comparable to motion-free scans in phantoms and healthy subjects, showing 10-20% improvement in image quality metrics (PSNR, SSIM, NRMSE).
• Significant reduction in motion artifacts was observed in pediatric and patient data.
• Demonstrated 20-30% reduction in N-acetylaspartate (NAA) linewidth and fitting error in a child.
• Achieved approximately 50% reduction in coefficient-of-variation in a patient with involuntary movements.

Conclusions:

• An effective motion correction technique for high-resolution, non-water-suppressed MRSI has been successfully developed.
• The proposed method shows significant potential to improve the reliability and clinical adoption of MRSI.
• This advancement facilitates more accurate metabolite quantification and robust neuroimaging in challenging patient populations.