Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Jul;44(10):4064-4076.
doi: 10.1002/hbm.26330. Epub 2023 May 5.

Distance-dependent distribution thresholding in probabilistic tractography

Ya-Ning Chang  1   2 , Ajay D Halai  2 , Matthew A Lambon Ralph  2
Affiliations

Distance-dependent distribution thresholding in probabilistic tractography

Ya-Ning Chang et al. Hum Brain Mapp. 2023 Jul.

Abstract

Tractography is widely used in human studies of connectivity with respect to every brain region, function, and is explored developmentally, in adulthood, ageing, and in disease. However, the core issue of how to systematically threshold, taking into account the inherent differences in connectivity values for different track lengths, and to do this in a comparable way across studies has not been solved. By utilising 54 healthy individuals' diffusion-weighted image data taken from HCP, this study adopted Monte Carlo derived distance-dependent distributions (DDDs) to generate distance-dependent thresholds with various levels of alpha for connections of varying lengths. As a test case, we applied the DDD approach to generate a language connectome. The resulting connectome showed both short- and long-distance structural connectivity in the close and distant regions as expected for the dorsal and ventral language pathways, consistent with the literature. The finding demonstrates that the DDD approach is feasible to generate data-driven DDDs for common thresholding and can be used for both individual and group thresholding. Critically, it offers a standard method that can be applied to various probabilistic tracking datasets.

Keywords: diffusion-weighted imaging; language connectome; probabilistic tractography; threshold selection.

PubMed Disclaimer

Figures

FIGURE 1
FIGURE 1
(a) The representative whole‐brain tracking with 10 M streamlines. (b) An example of the extracted streamlines connecting two ROIs in the inferior frontal cortex. (c) The 230 ROIs (a radius of 8 mm) covering the left hemisphere before overlapping with the grey/white matter interface.
FIGURE 2
FIGURE 2
The number of ROI paired samples for each of the 26 distance ranges based on Euclidean distance (a) and streamline distance (b).
FIGURE 3
FIGURE 3
The scatter plot of the distance scores based on Euclidean distance and streamline distance for all the ROI pairs.
FIGURE 4
FIGURE 4
The sampling distribution of connectivity for each of the 26 distance ranges based on the Euclidean distance (a) and streamline distance (b). The x‐axis indicates connection strength and the y‐axis indicates the number of samples on a log scale. DR: distance range (mm).
FIGURE 5
FIGURE 5
The distance‐dependent distribution thresholds at three alpha levels of 10%, 20% and 30% varied with the 26 ROI distance bins based on Euclidean distance (a) and streamline distance (b).
FIGURE 6
FIGURE 6
The average language connectivity matrix across individuals after thresholding and binarization based on Euclidean distance. The different levels of alpha with 10% on the top, 20% on the middle, and 30% on the bottom were superimposed onto one matrix wherein anything that survives the 10% would also survive 20% and 30%.
FIGURE 7
FIGURE 7
The projections of the average language connectivity matrix generated by using Euclidean distance in the brain with three alpha levels of 10% (red), 20% (green) and 30% (blue) and the difference plots between them (gold). The wider the connection line indicates the stronger the connections.
FIGURE 8
FIGURE 8
The average language connectivity matrix generated by using Euclidean distance with a common arbitrary thresholding approach at 5%, 10%, 20% and 40% of percentiles. The different levels of thresholding with 5% on the top, 10% and 20% on the middle, and 40% on the bottom were superimposed onto one matrix wherein anything that survives the 5% would also survive 10%, 20% and 40%.
See this image and copyright information in PMC

References

    1. Andersson, J. L. R. , Skare, S. , & Ashburner, J. (2003). How to correct susceptibility distortions in spin‐echo echo‐planar images: Application to diffusion tensor imaging. NeuroImage, 20(2), 870–888. 10.1016/S1053-8119(03)00336-7 - DOI - PubMed
    1. Andersson, J. L. R. , & Sotiropoulos, S. N. (2016). An integrated approach to correction for off‐resonance effects and subject movement in diffusion MR imaging. NeuroImage, 125, 1063–1078. 10.1016/j.neuroimage.2015.10.019 - DOI - PMC - PubMed
    1. Avants, B. B. , Tustison, N. , & Song, G. (2009). Advanced normalization tools (ANTS). Insight J, 2(365), 1–35.
    1. Behrens, T. E. J. , Berg, H. J. , Jbabdi, S. , Rushworth, M. F. S. , & Woolrich, M. W. (2007). Probabilistic diffusion tractography with multiple fibre orientations: What can we gain? NeuroImage, 34(1), 144–155. 10.1016/j.neuroimage.2006.09.018 - DOI - PMC - PubMed
    1. Behrens, T. E. J. , Woolrich, M. W. , Jenkinson, M. , Johansen‐Berg, H. , Nunes, R. G. , Clare, S. , Matthews, P. M. , Brady, J. M. , & Smith, S. M. (2003). Characterization and propagation of uncertainty in diffusion‐weighted MR imaging. Magnetic Resonance in Medicine, 50(5), 1077–1088. 10.1002/mrm.10609 - DOI - PubMed

Publication types