app-ensembletracking

This App uses MRtrix 0.2.12 to do ensemble tracking using tensor and constrained spherical deconvolution (csd) algorithms. It generates a large set of candidate streamlines using a tensor-based deterministic model, csd-based deterministic model, and csd-based probabilistic model. The csd-based models can be computed at lmax values of 2, 4, 6, 8, 10, and 12. All candidate streamlines are combined into a single track.mat file.

max lmax can be left blank and it will be calculated for you.

If you wish to use just deterministic tracking (MRtrix streamtrack parameter SD_STREAM) check do_deterministic.

If you wish to use just probabilistic tracking (MRtrix streamtrack parameter SD_PROB) check do_probabilistic.

If you wish to use the tensor tracking (MRtrix streamtrack parameter DT_STREAM) check do_tensor. By default it will use all three tracking methods.

For more information about Ensemble Tractography see Takemura, H., Caiafa, C. F., Wandell, B. A., & Pestilli, F. (2016). Ensemble tractography. PLoS computational biology, 12(2), e1004692.

Authors

• Lindsey Kitchell (kitchell@indiana.edu)
• Soichi Hayashi (hayashis@iu.edu)
• Brent McPherson (bcmcpher@iu.edu)

Project director

• Franco Pestilli (franpest@indiana.edu)

Funding

Running the App

On Brainlife.io

You can submit this App online

• with dtiInit input https://doi.org/10.25663/bl.app.33
• with dwi input https://doi.org/10.25663/bl.app.103

via the "Execute" tab.

Running Locally (on your machine)

1. git clone this repo.
2. Inside the cloned directory, create config.json with something like the following content with paths to your input files.

{
    "dwi": "/N/u/hayashis/Karst/testdata/108323/dwi/dwi.nii.gz",
    "bvecs": "/N/u/hayashis/Karst/testdata/108323/dwi/dwi.bvecs",
    "bvals": "/N/u/hayashis/Karst/testdata/108323/dwi/dwi.bvals",
    "freesurfer": "/N/u/hayashis/Karst/testdata/108323/freesurfer/output",
    "stepsize": 0.2,
    "minlength": 10,
    "maxlength": 200,
    "num_or_fibers": 0,
    "num_mt_fibers": 0,
    "num_vz_fibers": 0,
    "detr_curvs": "0.25 0.5 1 2 4",
    "prob_curvs": "0.25 0.5 1 2 4",
    "num_cc_fibers": 2500,
    "num_fibers": 12500,
    "do_tensor": true,
    "do_probabilistic": true,
    "do_deterministic": true
}

3. Launch the App by executing main

./main

Sample Datasets

If you don't have your own input file, you can download sample datasets from Brainlife.io, or you can use Brainlife CLI.

npm install -g brainlife
bl login
mkdir input
bl dataset download 5a050a00eec2b300611abff3 && mv 5a050a00eec2b300611abff3 input/dwi
bl dataset download 5a065cc75ab38300be518f51 && mv 5a065cc75ab38300be518f51 input/freesurfer

Output

track.tck

Dependencies

This App only requires singularity to run. If you don't have singularity, you will need to install following dependencies.

• Freesurfer 6
• FSL 5.0.9
• Mrtrix 0.2.12