Abstract: As tractography datasets continue to grow in size, there is a need for improved visualization methods that can capture structural patterns occurring in large tractography datasets. Transparency is an increasingly important aspect of finding these patterns in large datasets but is inaccessible to tractography due to performance limitations. In this paper, we propose a rendering method that achieves performant rendering of transparent streamlines, allowing for exploration of deeper brain structures interactively. The method achieves this through a novel approximate order-independent transparency method that utilizes voxelization and caching view-dependent line orders per voxel. We compare our transparency method with existing tractography visualization software in terms of performance and the ability to capture deeper structures in the dataset.
Example result
This video shows a dataset containing 140 thousand streamlines rendered in real-time using the transparency method described in the paper. It utilizes both voxelization and view-dependent line orders. For all examples, see the videos page.
BibTeX
@inproceedings{osman2023voxlines,
title={Voxlines: Streamline Transparency through Voxelization and View-Dependent Line Orders},
author={Besm Osman and Mestiez Pereira and Huub van de Wetering and Maxime Chamberland},
year={2023},
booktitle={Computational Diffusion MRI},
pages={92--103},
isbn={978-3-031-47292-3},
publisher={Springer Nature Switzerland}
}
Dataset
For our paper we generated a set of whole-brain tractograms from a single participant sourced from the Human Connectome Project. A tractogram consisting of one million streamlines was constructed using multi-shell multi-tissue constrained spherical deconvolution in MRtrix. Subsequently, a smaller dataset was created using TractSeg , resulting in a tractogram with 140k streamlines.
Download: whole_brain_140K.tck (44MB)
Download: whole_brain_1M.tck (659MB)