Related Experiment Video
Updated: Oct 23, 2025
17:06
Co-analysis of Brain Structure and Function using fMRI and Diffusion-weighted Imaging
Published on: November 8, 2012
26.5K
Tractography dissection variability: What happens when 42 groups dissect 14 white matter bundles on the same dataset?
Kurt G Schilling1, François Rheault2, Laurent Petit3
1Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States.
Neuroimage
|August 25, 2021
Summary
Diffusion MRI fiber tractography enables white matter bundle segmentation. However, diverse protocols create significant variability, impacting results and interpretation. Reducing this heterogeneity is crucial for reproducible research and clinical applications.
Area of Science:
- Neuroimaging
- Neuroscience
- Medical image analysis
Background:
- Diffusion MRI fiber tractography is key for in vivo human brain white matter pathway identification.
- Variability in segmentation protocols complicates tractography results, quantification, and interpretation.
Purpose of the Study:
- To evaluate and quantify the variability introduced by different white matter bundle segmentation protocols.
- To compare protocol variability against other sources of variability in tractography.
Main Methods:
- An open call distributed processed human whole-brain streamlines to 42 independent teams.
- Teams used 57 distinct protocols to segment 14 white matter fascicles across six subjects.
- Volume-based and streamline-based analyses assessed agreement and disagreement among protocols.
Main Results:
- Variability across bundle segmentation protocols exceeded other sources of variability, including within-protocol and inter-subject differences.
- This heterogeneity was observed even when using identical underlying streamline data.
Conclusions:
- Protocol variability is a major challenge in white matter bundle segmentation using diffusion MRI.
- Reducing this heterogeneity through standardized nomenclature and dissection methods is essential for reproducible research and clinical translation.
- External validation is needed to confirm anatomical accuracy of segmented bundles.
