BundleParc: Consistent white matter bundle parcellation without tractography

Area of Science:

• Neuroimaging
• White Matter Anatomy
• Computational Neuroscience

Background:

• Tractometry analyzes microstructural properties of white matter (WM) tracts.
• Accurate white matter bundle parcellation is crucial for tractometry.
• Current methods are inconsistent, computationally expensive, and lack reproducibility.

Purpose of the Study:

• Introduce BundleParc, a novel, tractography-free bundle parcellation method.
• Develop a consistent, accurate, and reproducible parcellation technique.
• Simplify and enhance the utility of white matter bundle segmentation.

Main Methods:

• BundleParc directly maps fiber orientation distribution function (fODF) volumes to label maps.
• Avoids computationally intensive streamline generation and tractography.
• Utilizes a direct mapping approach for parcellation.

Main Results:

• BundleParc demonstrates superior accuracy, robustness, and reproducibility compared to state-of-the-art methods.
• The method is significantly simpler and faster than existing tract-based profiling techniques.
• Evaluations on research and clinical cohorts confirm its effectiveness.

Conclusions:

• BundleParc provides a consistent and tractography-free solution for bundle parcellation.
• It offers a significant advancement for reproducible and efficient white matter analysis.
• The method is suitable for fast, easy-to-use, off-the-shelf bundle segmentation.