3D curve inference for diffusion MRI regularization.

Area of Science:

• Medical Imaging
• Computational Neuroscience
• Differential Geometry

Background:

• Diffusion MRI is crucial for mapping white matter tracts.
• Existing regularization methods for diffusion MRI face limitations in accuracy and consistency.
• Accurate fiber orientation estimation is essential for understanding brain connectivity.

Purpose of the Study:

• To develop a novel differential geometric framework for regularizing diffusion MRI data.
• To improve the accuracy and robustness of white matter fiber orientation estimation.
• To offer advantages over existing regularization techniques in diffusion MRI analysis.

Main Methods:

• Modeling white matter fibers as 3D space curves.
• Extending a 2D curve inference approach using co-helicity.
• Applying the framework to diffusion MRI data regularization.

Main Results:

• Demonstrated quantitative validation on biological phantom and synthetic data.
• Showcased qualitative validation on in vivo human brain data.
• The co-helicity approach effectively indicates compatibility between neighboring fiber orientations.

Conclusions:

• The developed differential geometric framework provides effective regularization for diffusion MRI data.
• This method offers advantages over previous regularization techniques.
• The approach shows promise for enhanced white matter tractography and brain connectivity studies.