Incorporating outlier information into diffusion-weighted MRI modeling for robust microstructural imaging and...

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Medical Physics

Background:

• Diffusion-weighted MRI tractography visualizes white matter but suffers from false positives, limiting brain connectivity analysis.
• Existing filtering methods fail with motion artifacts common in clinical data, particularly in vulnerable populations like infants.

Purpose of the Study:

• To augment the COMMIT algorithm to robustly filter tractography data affected by motion-induced signal dropouts.
• To improve the specificity and accuracy of structural brain connectivity analyses in the presence of motion artifacts.

Main Methods:

• Augmented the Convex Optimization Modelling for Microstructure Informed Tractography (COMMIT) algorithm to handle motion artifacts.
• Validated the enhanced algorithm using comprehensive Monte-Carlo whole-brain simulations and in vivo infant MRI data.

Main Results:

• The robust COMMIT augmentation effectively filtered tractography reconstructions despite motion artifacts.
• Motion artifacts significantly and adversely affect human brain structural connectivity and microstructural property mapping if not addressed.

Conclusions:

• Robust filtering methods are crucial for mitigating motion-related errors in tractogram filtering, especially in pediatric clinical studies.
• The presented robust COMMIT augmentation offers readily available, open-source solutions for improved tractography analysis.