Reducing Susceptibility Distortion Related Image Blurring in Diffusion MRI EPI Data.

Area of Science:

• Neuroscience
• Biomedical Imaging
• Computational Neuroscience

Background:

• Diffusion MRI is crucial for neuroscience, enabling myelin mapping via g-ratio analysis.
• Statistical Parametric Mapping (SPM) is a key tool for g-ratio group analyses.
• Existing SPM methods for susceptibility distortion correction in diffusion MRI have limitations, causing image blurring.

Purpose of the Study:

• To enhance susceptibility-related distortion correction for diffusion MRI data within SPM.
• To implement an alternative to the arithmetic mean (AM) for reducing blurring in diffusion MRI data.
• To evaluate the impact of these improvements on g-ratio mapping accuracy.

Main Methods:

• Developed and implemented a new processing pipeline: consecutive Hyperelastic Susceptibility Artefact Correction (HySCO).
• Employed a weighted average (WA) method to combine blip-up and blip-down echo-planar imaging (EPI) data.
• Assessed improvements using diffusion MRI, multi-parameter mapping (MPM), and g-ratio mapping.

Main Results:

• The consecutive HySCO pipeline demonstrated superior distortion correction compared to standard ACID toolbox methods.
• The WA method resulted in better alignment of diffusion data with structural white matter maps.
• Consecutive HySCO with WA outperformed the AM method for g-ratio mapping.

Conclusions:

• The improved HySCO-WA pipeline enhances susceptibility distortion correction and reduces blurring in diffusion MRI data.
• These advancements improve the accuracy and anatomical fidelity of g-ratio mapping.
• The open-source ACID toolbox in SPM now offers improved diffusion MRI analysis, facilitating myelin biomarker computation.