Skip to content
Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


Alt text

Tool for fast and accurate white matter bundle segmentation from Diffusion MRI. It can create bundle segmentations, segmentations of the endregions of bundles and Tract Orientation Maps (TOMs). Moreover, it can do tracking on the TOMs creating bundle-specific tractogram and do Tractometry analysis on those.

The tool works very well for data similar to the Human Connectome Project. For other MRI datasets it works well for all bundles except for the Commissure Anterior (CA) and the Fornix (FX) which are incomplete sometimes.

TractSeg is the code for the following papers. Please cite the papers if you use it.

Build Status Brainlife

Table of contents


TractSeg only runs on Linux and OSX. It works with Python >= 3.6.

Install Prerequisites

  • Pytorch
  • Mrtrix 3 (>= 3.0 RC3)
  • FSL (if you already have a brain mask and do not use the option --preprocess this is not needed)

Install TractSeg

Latest stable version:

pip install TractSeg

NOTE: See CHANGELOG for (breaking) changes of each version


You can also directly use TractSeg via Docker (contains all prerequisites).

sudo docker run -v /absolute/path/to/my/data/directory:/data \
-t wasserth/tractseg_container:master TractSeg -i /data/my_diffusion_file.nii.gz -o /data --raw_diffusion_input

On OSX you might have to increase the Docker memory limit from the default of 2GB to something like 7GB.

How to use

Simple example:

To segment the bundles on a Diffusion Nifti image run the following command. (Diffusion.bvals and Diffusion.bvecs have to be in the same directory as the input image.) (You can use the example image provided in this repository under examples.)

TractSeg -i Diffusion.nii.gz --raw_diffusion_input

This will create a folder tractseg_ouput inside of the same directory as your input file with one binary segmentation nifti image for each bundle.

NOTE: Your input image should have the same orientation as MNI space. Moreover it should have isotropic spacing. See here for more details.

Custom input and output path:

TractSeg -i my/path/my_diffusion_image.nii.gz
         -o my/output/directory
         --bvals my/other/path/my.bvals
         --bvecs yet/another/path/my.bvec

Use --help to see all options.

Use existing peaks

To avoid generating the MRtrix CSD peaks every time you run TractSeg you can also provide them directly by skipping the option --raw_diffusion_input`.

TractSeg -i my/path/my_mrtrix_csd_peaks.nii.gz

Segment bundle start and end regions

Get segmentations of the regions were the bundles start and end (helpful for filtering fibers that do not run from start until end).

TractSeg -i peaks.nii.gz --output_type endings_segmentation

Create Tract Orientation Maps (TOMs)

For each bundle create a Tract Orientation Map (Wasserthal et al., Tract orientation mapping for bundle-specific tractography). This gives you one peak per voxel telling you the main orientation of the respective bundle at this voxel. Can be used for bundle-specific tracking later on.

TractSeg -i peaks.nii.gz --output_type TOM

NOTE: --output_type tract_segmentation and endings_segmentation has to be run first. Same input and output directories have to be used for all commands.

Peaks and streamlines can be visualized using for example MITK Diffusion.

NOTE: Peaks have to be flipped along the z-axis to be displayed correctly in MITK.

Create bundle-specific tractograms

Tracks on TOMs and only keeps fibers not leaving the bundle mask and starting and ending in the endpoint regions.

TractSeg -i peaks.nii.gz --output_type tract_segmentation
TractSeg -i peaks.nii.gz --output_type endings_segmentation
TractSeg -i peaks.nii.gz --output_type TOM 
Tracking -i peaks.nii.gz

NOTE: If you are not using MITK Diffusion for viewing your results you might want to use a different tracking format by adapting the option --tracking_format.

Use bedpostX peaks instead of CSD peaks

TractSeg also works with bedpostX as input. You have to pass dyads1.nii.gz as input and TractSeg will automatically find all the other necessary bedpostX files (dyads2_thr0.05.nii.gz & dyads3_thr0.05.nii.gz. mean_f1-3samples will be used for scaling the peaks.). This only works if you did not change the default bedpostX file naming.

TractSeg -i dyads1.nii.gz

Show uncertainty map

Create map showing where the method is uncertain about its segmentation (uses monte carlo dropout:

TractSeg -i peaks.nii.gz --uncertainty

Perform Tractometry

See Documentation of Tractometry.

Extended Tutorial

Best pratices for standard usecases.
(You can also check out the tutorial from OHBM 2019 [OUTDATED].)

Track subset of bundles

You can specify to only track a subset of bundles.

Tracking -i peaks.nii.gz --bundles CST_right,CA,IFO_right

Bundle names

The following list shows the index of each extracted bundle in the output file (if using --single_output_file).

0: AF_left         (Arcuate fascicle)
1: AF_right
2: ATR_left        (Anterior Thalamic Radiation)
3: ATR_right
4: CA              (Commissure Anterior)
5: CC_1            (Rostrum)
6: CC_2            (Genu)
7: CC_3            (Rostral body (Premotor))
8: CC_4            (Anterior midbody (Primary Motor))
9: CC_5            (Posterior midbody (Primary Somatosensory))
10: CC_6           (Isthmus)
11: CC_7           (Splenium)
12: CG_left        (Cingulum left)
13: CG_right   
14: CST_left       (Corticospinal tract)
15: CST_right 
16: MLF_left       (Middle longitudinal fascicle)
17: MLF_right
18: FPT_left       (Fronto-pontine tract)
19: FPT_right 
20: FX_left        (Fornix)
21: FX_right
22: ICP_left       (Inferior cerebellar peduncle)
23: ICP_right 
24: IFO_left       (Inferior occipito-frontal fascicle) 
25: IFO_right
26: ILF_left       (Inferior longitudinal fascicle) 
27: ILF_right 
28: MCP            (Middle cerebellar peduncle)
29: OR_left        (Optic radiation) 
30: OR_right
31: POPT_left      (Parieto‐occipital pontine)
32: POPT_right 
33: SCP_left       (Superior cerebellar peduncle)
34: SCP_right 
35: SLF_I_left     (Superior longitudinal fascicle I)
36: SLF_I_right 
37: SLF_II_left    (Superior longitudinal fascicle II)
38: SLF_II_right
39: SLF_III_left   (Superior longitudinal fascicle III)
40: SLF_III_right 
41: STR_left       (Superior Thalamic Radiation)
42: STR_right 
43: UF_left        (Uncinate fascicle) 
44: UF_right 
45: CC             (Corpus Callosum - all)
46: T_PREF_left    (Thalamo-prefrontal)
47: T_PREF_right 
48: T_PREM_left    (Thalamo-premotor)
49: T_PREM_right 
50: T_PREC_left    (Thalamo-precentral)
51: T_PREC_right 
52: T_POSTC_left   (Thalamo-postcentral)
53: T_POSTC_right 
54: T_PAR_left     (Thalamo-parietal)
55: T_PAR_right 
56: T_OCC_left     (Thalamo-occipital)
57: T_OCC_right 
58: ST_FO_left     (Striato-fronto-orbital)
59: ST_FO_right 
60: ST_PREF_left   (Striato-prefrontal)
61: ST_PREF_right 
62: ST_PREM_left   (Striato-premotor)
63: ST_PREM_right 
64: ST_PREC_left   (Striato-precentral)
65: ST_PREC_right 
66: ST_POSTC_left  (Striato-postcentral)
67: ST_POSTC_right
68: ST_PAR_left    (Striato-parietal)
69: ST_PAR_right 
70: ST_OCC_left    (Striato-occipital)
71: ST_OCC_right

Use different tract definitions

TractSeg was also trained on the bundles provided by XTRACT. These bundles are slightly differently defined. They tend to be more specific but therefore also a bit less complete. Depending on your application this might be of interest for you. Using TractSeg instead of XTRACT has the advantage of being at least 10x faster. Please cite XTRACT if you use this.

NOTE: This is only supported for output type tract_segmentation and dm_regression.

TractSeg -i peaks.nii.gz --tract_definition xtract

Use python interface

import nibabel as nib
import numpy as np
from tractseg.python_api import run_tractseg
peaks = nib.load("tests/reference_files/peaks.nii.gz").get_fdata()
segmentation = run_tractseg(peaks)

Different tracking types

You can use different types of tracking:

  • "Probabilistic" tracking on TOM peaks [default].
    Tracking -i peaks.nii.gz
    Probabilistic means that at each step a small random factor will be added to the direction given by the TOM peaks. If not doing this on low resolution data it sometimes gets difficult finding fibers running from start to end and covering the whole bundle.

  • Probabilistic tracking on original FODs.
    Tracking -i WM_FODs.nii.gz --track_FODs iFOD2
    Is calling Mrtrix iFOD2 tracking internally. Does not use TOM peaks but the original FODs. The results will get filtered by the bundle mask and have to start and end in the endings masks.

Tracking formats

You can use the option --tracking_format to define the file format of the streamline files. The default is trk_legacy. The resulting streamlines are displayed correctly in MITK Diffusion. If you are using TrackVis you should use the format trk.

NOTE: When calling Tractometry you have to set the same tracking format as was used in Tracking.

Aligning image to MNI space

The input image must have the same "orientation" as the Human Connectome Project data (MNI space) (LEFT must be on the same side as LEFT of the HCP data) and have isotropic spacing. If the image orientation and the gradient orientation of your data is the same as in examples/Diffusion.nii.gz you are fine. Otherwise you should use fslreorient2std or rigidly register your image to MNI space (the brains do not have to be perfectly aligned but must have the same LEFT/RIGHT orientation). You can use the following FSL commands to rigidly register you image to MNI space (uses the FA to calculate the transformation as this is more stable):

calc_FA -i Diffusion.nii.gz -o FA.nii.gz --bvals Diffusion.bvals --bvecs Diffusion.bvecs \
--brain_mask nodif_brain_mask.nii.gz

flirt -ref tractseg/tractseg/resources/MNI_FA_template.nii.gz -in FA.nii.gz \
-out FA_MNI.nii.gz -omat FA_2_MNI.mat -dof 6 -cost mutualinfo -searchcost mutualinfo

flirt -ref tractseg/tractseg/resources/MNI_FA_template.nii.gz -in Diffusion.nii.gz \
-out Diffusion_MNI.nii.gz -applyxfm -init FA_2_MNI.mat -dof 6
cp Diffusion.bvals Diffusion_MNI.bvals
rotate_bvecs -i Diffusion.bvecs -t FA_2_MNI.mat -o Diffusion_MNI.bvecs

To enforce isotropic spacing you can replace -applyxfm by -applyisoxfm <your_spacing>.

To move the results back to subject space you can use the following commands:

convert_xfm -omat MNI_2_FA.mat -inverse FA_2_MNI.mat  # invert transformation

flirt -ref FA.nii.gz -in my_bundle.nii.gz -out my_bundle_subject_space.nii.gz \
-applyxfm -init MNI_2_FA.mat -dof 6 -interp spline  # for TOM maps you have to use the command vecreg

fslmaths my_bundle_subject_space.nii.gz -thr 0.5 -bin my_bundle_subject_space.nii.gz  # float to binary 

The option --preprocess will automatically rigidly register the input image to MNI space, run TractSeg and then convert the output back to subject space. For TOMs and trackings the --preprocess option has to be used as follows:

# in first step --raw_diffusion_input has to be used together with --preprocess
TractSeg -i Diffusion.nii.gz -o tractseg_output --output_type tract_segmentation --raw_diffusion_input --preprocess
# -o has to be set in all following steps
TractSeg -i tractseg_output/peaks.nii.gz -o tractseg_output --output_type endings_segmentation --preprocess
TractSeg -i tractseg_output/peaks.nii.gz -o tractseg_output --output_type TOM --preprocess
Tracking -i tractseg_output/peaks.nii.gz -o tractseg_output

NOTE: --preprocess does not work if you are using the option --csd_type csd_msmt_5tt, because the T1 image will not automatically be registered to MNI space

If you are not familiar with preprocessing of DWI images (e.g. correcting for artifacts) you can have a look at this preprocessing script.


My output segmentation does not look like any bundle at all!

Make sure your input image is in MNI space. Even if the input image is in MNI space the Mrtrix peaks might still be flipped. In those cases you should view the peaks in mrview and make sure they have the proper orientation. Otherwise you might have to flip the sign along the x, y or z axis using the following command:

flip_peaks -i my_peaks.nii.gz -o my_peaks_flip_y.nii.gz -a y

My data has anisotropic spacing. Does it still work?

If the spacing is only slightly anisotropic (e.g. 1.9mm x 1.9mm x 2mm) it still works fine. If the spacing is heavily anisotropic (e.g. 1mm x 1mm x 2mm) results will be a lot better if you resample your image to isotropic spacing first.

Small bundles like the CA and FX are incomplete

You can use the following options to improve your results:
--super_resolution The input image is upsampled to 1.25mm resolution (the resolution TractSeg was trained on) and finally downsampled back to the original resolution. Using --super_resolution will output the image at 1.25mm. Especially if image resolution is low parts of the CA can get lost during downsampling.

Can I save the pretrained weights at a different location?

Per default the pretrained weights will be downloaded to and loaded from ~/.tractseg/. You can change this directory by adding weights_dir=/absolute/path_to_where/you_want_it to ~/.tractseg/config.txt in a new line (if the file does not exist yet you have to create it).
Normally the pretrained weights will automatically be downloaded in the background right when they are needed. In some cases you might want to download all of them at once. To do so you can simply run download_all_pretrained_weights and the weights will be download to ~/.tractseg/ or the location you specified in ~/.tractseg/config.txt.

Did I install the prerequisites correctly?

You can check if you installed Mrtrix correctly if you can run the following command on your terminal: dwi2response -help

You can check if you installed FSL correctly if you can run the following command on your terminal: bet -help

TractSeg uses these commands so they have to be available.

How can I install the latest master branch?

pip install

Train your own model

TractSeg uses a pretrained model. You can also train your own model on your own data, however TractSeg is not optimised to make this convenient, as most people will use the pretrained model. The following guide is quite short and you might have problems following every step. Contact the author if you need help training your own model.

  1. Use your own data or download the data from Zenodo
  2. If you have streamlines you have to transform them to binary masks. You can use this script for that. (If you want to train a TOM model you have to create peak maps from the streamlines. This can be done by using the MITK Diffusion miniapp MitkFiberDirectionExtraction. See the following two issues for more details: 82, 92)
  3. Install TractSeg from local source code:
    git clone
    pip install -e TractSeg
  4. Install BatchGenerators:
    git clone
    pip intall -e batchgenerators
  5. The folder structure of your training data should be the following:
          '-> mrtrix_peaks.nii.gz       (mrtrix CSD peaks;  shape: [x,y,z,9])
          '-> bundle_masks.nii.gz       (Reference bundle masks; shape: [x,y,z,nr_bundles])
  6. Preprocess the data using tractseg/data/ to remove all non-brain area (crop to brain bounding box). Adapt the data pathes in tractseg/data/ to fit your data (look for #todo: adapt inside of the file.)
  7. Adapt the file tractseg/config/custom/
  8. Create a file ~/.tractseg/config.txt. This contains the path to your data directory working_dir=XXX, e.g. working_dir=custom_path.
  9. Adapt with the bundles you use in your reference data.
  10. Adapt with the names of your label files.
  11. Adapt with the list of your subject IDs.
  12. Run ExpRunner --config My_custom_experiment
  13. custom_path/hcp_exp/My_custom_experiment contains the results


To build a docker container with all dependencies run the following command in project root:

sudo docker build -t tractseg_container -f Dockerfile_CPU .


Copyright © German Cancer Research Center (DKFZ), Division of Medical Image Computing (MIC). Please make sure that your usage of this code is in compliance with the code license.