Tractome (http://tractome.org) is an interactive clustering-based 3D tool for exploration and segmentation of tractography data.
Tractome is an interactive tool for visualisation, exploration and segmentation of tractography data. It supports neuroanatomists and medical doctors in their study of white matter anatomical structures from diffusion magnetic resonance imaging (dMRI) data. Unlike previous systems for tractography segmentation, Tractome is a computer-assisted tool in which the user interacts with a summary of the tractography instead of the whole set of streamlines. The summary is generated by clustering the tractography into a desired number of clusters, usually in the order of tens or one hundred. The summary shows only one representative streamline for each cluster, so that it is easier to interact with them in the 3D scene. The user can then iteratively select the representative streamlines of interest and re-cluster the associated sets or streamlines into smaller ones in order to incrementally reveal details of the anatomical structure of interest. The interaction is powered by novel efficient algorithms for fast clustering. Tractome is written in Python language and it supports the TrackVis format and the DiPy format for tractography files.
To install the tractome from the source, we need to install the dependencies. Execute the following command in the project root directory.
pip install -r requirements.txt
Once, we gather the dependencies we can install the tractome locally. (Required: If you want to use the CLI).
pip install -e .
To run the tractome
tractome --tractogram <file_path> --mesh <file_path> --mesh_tex <file_path> --t1 <file_path>
NOTE: All the options above are optional you can start and empty window as well. The functionality to add/swap files will be added later in the UI.
To visualize cluster representation
You are required to pre-compute the embeddings of the streamlines, use the following command to save a new .trx file with embedding. Later you can provide this newly saved file to open the tractogram in the above command.
tractome_compute_dis_matrix <file_path>
For getting a complete list and details of the options available.
tractome_compute_dis_matrix --help