Motion correction and registration of high b-value diffusion weighted images

Area of Science:

• Biomedical Imaging
• Neuroimaging
• Diffusion MRI

Background:

• High b-value diffusion weighted MRI enhances sensitivity and specificity to tissue microstructure compared to clinical b-values.
• However, high b-value diffusion MRI is prone to poor signal-to-noise ratio, necessitating longer acquisition times and increasing motion artifacts.
• The orientational sensitivity and varying contrast at different b-values and gradient directions complicate conventional motion correction methods.

Purpose of the Study:

• To develop and validate a framework for correcting motion-induced misalignments and artifacts in high b-value diffusion weighted MRI.
• To improve the accuracy and reliability of diffusion MRI data acquired at high b-values.

Main Methods:

• A novel registration framework combining experimental diffusion tensor MRI data and simulations using the composite hindered and restricted model of diffusion (CHARMD) was proposed.
• The framework was evaluated using visual assessment of registered images and CHARMD analysis results.
• Residual analysis was employed to quantify the quality of CHARMD fitting.

Main Results:

• The proposed registration framework demonstrated improved fitting of diffusion MRI data to the CHARMD model.
• Both qualitative and quantitative assessments confirmed the effectiveness of the method in correcting motion and distortions.
• The results indicate a significant enhancement in data quality and model fitting.

Conclusions:

• The developed framework successfully addresses the challenge of motion and distortion correction in high b-value diffusion weighted MRI.
• This approach makes the correction of artifacts in high b-value diffusion MRI data feasible for the first time.
• The findings pave the way for more accurate and reliable diffusion MRI studies utilizing high b-values.