Installation

You will find below the steps for installing clinicadl on Linux or Mac. Please do not hesitate to contact us on the forum or GitHub if you encounter any issues.

Prepare your Python environment

You will need a Python environment to run ClinicaDL. We advise you to use Miniconda. Miniconda allows you to install, run, and update Python packages and their dependencies. It can also create environments to isolate your libraries. To install Miniconda, open a new terminal and type the following commands:

• If you are on Linux:

  curl https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -o /tmp/miniconda-installer.sh
  bash /tmp/miniconda-installer.sh
  

• If you are on Mac:

  curl https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh -o /tmp/miniconda-installer.sh
  bash /tmp/miniconda-installer.sh
  

Install ClinicaDL

The latest release of Clinica can be installed using pip as follows:

conda create --name clinicadlEnv python=3.7
conda activate clinicadlEnv
pip install clinicadl

Run the ClinicaDL environment

Activation of the ClinicaDL environment

Now that you have created the ClinicaDL environment, you can activate it:

conda activate clinicadlEnv

Success

Congratulations, you have installed ClinicaDL! At this point, you can try the basic clinicadl -h command and get the help screen:

(clinicadlEnv)$ clinicadl -h
usage: clinicadl [-h] [--verbose]
                 {generate,preprocessing,extract,train,classify,tsvtool} ...

Deep learning software for neuroimaging datasets

optional arguments:
  -h, --help            show this help message and exit
  --verbose, -v

Task to execute with clinicadl::
  What kind of task do you want to use with clinicadl? (tsvtool,
  preprocessing, extract, generate, train, validate, classify).

  {generate,preprocessing,extract,quality_check,train,classify,tsvtool}
                        ****** Tasks proposed by clinicadl ******
    generate            Generate synthetic data for functional tests.
    preprocessing       Prepare data for training (needs clinica installed).
    extract             Create data (slices or patches) for training.
    quality_check       Performs quality check procedure for t1-linear
                        pipeline.Original code can be found at
                        https://github.com/vfonov/deep-qc
    train               Train with your data and create a model.
    classify            Classify one image or a list of images with your
                        previously trained model.
    tsvtool             Handle tsv files for metadata processing and data
                        splits

Deactivation of the ClinicaDL environment

At the end of your session, remember to deactivate your Conda environment:

conda deactivate

Test ClinicaDL

Warning

Data for testing is not currently provided, but release of anonymized datasets for testing is planned for future versions.