TractSeg - Fast and accurate white matter tract segmentation

Abstract

The individual course of white matter fiber tracts is an important factor for analysis of white matter characteristics in healthy and diseased brains. Diffusion-weighted MRI tractography in combination with region-based or clustering-based selection of streamlines is a unique combination of tools which enables the in-vivo delineation and analysis of anatomically well-known tracts. This, however, currently requires complex, computationally intensive processing pipelines which take a lot of time to set up. TractSeg is a novel convolutional neural network-based approach that directly segments tracts in the field of fiber orientation distribution function (fODF) peaks without using tractography, image registration or parcellation. We demonstrate that the proposed approach is much faster than existing methods while providing unprecedented accuracy, using a population of 105 subjects from the Human Connectome Project. We also show initial evidence that TractSeg is able to generalize to differently acquired data sets for most of the bundles. The code and data are openly available at https://github.com/MIC-DKFZ/TractSeg/ and https://doi.org/10.5281/zenodo.1088277, respectively.

Tractostorm: The what, why, and how of tractography dissection reproducibility

“…Automatic segmentation methods are becoming more widespread. Methods such as, but not limited to, (Chekir, Descoteaux, Garyfallidis, Côté, & Boumghar, 2014;Garyfallidis et al, 2017;Guevara et al, 2011;O'Donnell et al, 2017;O'donnell, Golby, & Westin, 2013;Wassermann et al, 2016;Wasserthal, Neher, & Maier-Hein, 2018;Yendiki et al, 2011;Zhang et al, 2018) aim to simplify the work of raters. The typical standard of most automatic segmentation method is to reach the accuracy of raters, thus it is crucial to truly quantify human reproducibility in manual tasks.…”

Impact of b-value on estimates of apparent fibre density

“…In order to estimate AFD along various commonly investigated white matter fibre pathways, white matter tract segmentation was performed. We applied the automated TractSeg technique (Wasserthal, et al, 2018;Wasserthal, et al, 2019) in population template space, as this technique provides a balance between manual dissection and atlas-based tracking approaches. Of the existing library of 72 tracts, we chose to delineate 38 commonly investigated fibre pathways bilaterally for the left (L) and right (R) hemisphere ( Figure S2).…”

Evaluating the performance of 3-tissue constrained spherical deconvolution pipelines for within-tumor tractography

“…False positive streamlines and tracts are however of greater concern to certain non-targeted (automated) whole-brain analysis techniques, such as typical connectomics pipelines. Different approaches are being proposed to tackle this long-standing challenge, including machine learning techniques that are pretrained with a comprehensive set of known (anatomically valid) WM tracts (Wasserthal et al, 2018) and model-driven strategies which try to explain the data using a sparse set of tracts and other priors (Schiavi et al, 2019). Improvements in the reconstruction of within-tumor WM FODs, such as those achieved by SS3T-CSD, can directly provide a more reliable starting point for these advanced tractography strategies.…”

“…Although the challenges related to tumor infiltration thus might be effectively addressed, tumor mass effects on the other hand might prove to pose unique and non-trivial challenges to pretrained machine learning strategies, such as (Wasserthal et al, 2018) in particular, as these often partially rely on an expected location, shape or size of specific WM bundles. While beyond the scope of our work, in-depth evaluation of what kinds of features certain techniques-including more complex machine learning strategies-rely upon to perform robustly is an interesting avenue for future research.…”