Installation¶

DCEPrep can be run via Docker (recommended) or installed directly on your system.

Docker (Recommended)¶

Docker bundles all dependencies — FSL, ANTs, FreeSurfer, MATLAB, ROCKETSHIP, and the Python environment — into a single ~18 GB image. This is the easiest and most reproducible way to run DCEPrep.

Prerequisites¶

• Docker installed and running
• A valid MATLAB license file
• A valid FreeSurfer license file (license.txt)
• Your BIDS data directory accessible on the host

1. Clone the repository¶

git clone https://github.com/petmri/DCEPrep.git
cd DCEPrep

The main branch is stable. For a pinned release, check out a specific tag:

git checkout v1.0.0

2. Configure run_docker.sh¶

Open run_docker.sh and edit line 24 to set the path to your data directory on the host machine.

3. Run¶

./run_docker.sh

This script will pull the image if it isn't already cached and launch a container with the correct mounts.

Available image tags¶

Images are available on Docker Hub for multiple MATLAB releases:

docker pull lsaca05/dce:<MATLAB_release>-<branch>

# Examples:
docker pull lsaca05/dce:R2023a-main
docker pull lsaca05/dce:R2022a-dev

See Docker Hub for all available tags.

Without Docker¶

If you prefer to run DCEPrep directly without a container, you must install all dependencies manually.

Python environment¶

cd DCEPrep
python3 -m venv tf
source tf/bin/activate
pip install -r venv_requirements.txt

conda is also supported — install packages from venv_requirements.txt into a conda environment.

System dependencies¶

The following tools must be installed and available in your PATH:

Next Steps¶

• Set up your BIDS data
• Process Data guide