TBSS repository is developed by Tashrif Billah, Sylvain Bouix, and Ofer Pasternak, Brigham and Women's Hospital (Harvard Medical School).
If this repository is useful in your research, please cite as below:
Billah, Tashrif; Bouix, Sylvain; Pasternak, Ofer; Generalized Tract Based Spatial Statistics (TBSS) pipeline, https://github.com/pnlbwh/tbss, 2019, DOI: https://doi.org/10.5281/zenodo.2662497
See documentation for details.
Table of Contents created by gh-md-toc
- ANTs = 2.2.0
- FSL = 5.0.11
- numpy = 1.16.2
- pandas = 1.2.1
- dipy = 0.16.0
- nibabel = 2.3.0
- nilearn = 0.5.2
- pynrrd = 0.3.6
- conversion = 2.0
NOTE The above versions were used for developing the repository. However, tbss should work on any advanced version.
You may ignore installation instruction for any software module that you have already.
uname -a # check if 32 or 64 bit
Download Miniconda Python 3.6 bash installer (32/64-bit based on your environment):
sh Miniconda3-latest-Linux-x86_64.sh -b # -b flag is for license agreement
Activate the conda environment:
source ~/miniconda3/bin/activate # should introduce '(base)' in front of each line
Follow the instruction to download and install FSL.
(Preferred) You should install pre-complied ANTs from PNL-BWH:
conda install -c pnlbwh ants
Installation with conda is more manageable. It will put the ANTs commands/scripts in your path when you do:
source ~/miniconda3/bin/activate
Alternatively, you can build ANTs from source.
Now that you have installed the prerequisite software, you are ready to install the pipeline:
git clone https://github.com/pnlbwh/tbss && cd tbss
./install setup test
If you would not like to run tests, just omit the test argument. But it is recommended to run tests before you use
this pipeline to analyze your data.
Make sure the following executables are in your path:
antsMultivariateTemplateConstruction2.sh
antsApplyTransforms
antsRegistrationSyNQuick.sh
tbss_1_preproc
You can check them as follows:
which tbss_1_preproc
If any of them does not exist, add that to your path:
export PATH=$PATH:/directory/of/executable
ANTs commands should be in ~/miniconda3/pkgs/ants-2.3.0-py3/bin and/or ~/miniconda3/bin directories.
If they are not in your path already, use export PATH=$PATH:~/miniconda3/pkgs/ants-2.3.0-py3/bin
to put all the commands in your path. Additionally, you should define define ANTSPATH:
export ANTSPATH=~/miniconda3/bin
Upon successful installation, you should be able to see the help message
$ lib/tbss-pnl --help
See Useful commands for quick tips about running the pipeline.
Test includes both pipeline test and unit tests. It is recommended to run tests before analyzing your data.
The repository comes with separate testing for three branches: --enigma, --fmrib, and --studyTemplate:
./install.sh test
Running the tests should take less than an hour.
You may run smaller and faster unit tests as follows.
pytest -v lib/tests/test_*.py
NOTE In the current release, unit tests are dependant upon the outputs of whole pipeline test. This is likely to change in future.