Topographic Filtering of Tractograms as Vector Field Flows

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biophysics

Background:

• Diffusion MRI (dMRI) is a key tool for in vivo brain connectivity studies.
• The biological accuracy of dMRI-based tractography remains a challenge.
• Topographic regularity in brain circuits offers a way to validate tractography.

Purpose of the Study:

• To develop a novel framework for analyzing topographic regularity in dMRI tractography.
• To improve the reliability of brain connectivity analysis using diffusion MRI data.
• To filter out biologically implausible fiber tracts.

Main Methods:

• Developed a framework analyzing topographic regularity of brain connectivity from dMRI.
• Computed a principal vector field (PVF) from fiber orientation distributions (FODs) using Markov Random Fields.
• Quantified tract consistency with PVF using a Vector Flow Deviation (VFD) measure to filter irregular tracts.

Main Results:

• Successfully applied the VFD measure to filter irregular fiber tracts in the visual and colossal motor pathways.
• Reconstructed fiber bundles from Human Connectome Project (HCP) multi-shell dMRI data.
• Demonstrated filtering effectiveness using quantitative evaluation with G²SD distance.

Conclusions:

• The proposed method effectively filters topographically irregular fiber tracts, enhancing dMRI tractography reliability.
• This framework leverages mathematical properties of vector fields to improve biological validity of brain connectivity maps.
• The approach offers a promising tool for more accurate in vivo neuroscience research.