GirderEGGAnnotator

Works only on Linux/Mac, not on Windows.

Create environment and install dependencies

python -m venv .venv
source env/bin/activate
pip install .

Contributing

Install development dependencies

pip install -e ".[dev]"

Optional dependencies

---

Faster Jpeg encoding using TurboJPEG.

macOS system install

# macOS
brew install jpeg-turbo

Windows install

Download and install from GitHub: https://github.com/libjpeg-turbo/libjpeg-turbo/releases

Linux install

# RHEL/CentOS/Fedora
# YUM doc: https://libjpeg-turbo.org/Downloads/YUM
# Ubuntu
apt-get install libturbojpeg

Once your system is ready, you can install the dependencies:

pip install ".[turbo]"

Run trame application

Configuration file

Copy the config.template.yaml file to a config.yaml file that will be read by the app and fill in the configuration. This configuration file holds the style of the application and the configuration of the Girder backend. This application expects Girder running with the GirderBIDS plugin available here

# Clone the library
cd ..
git clone https://github.com/KitwareMedical/GirderBIDS.git
cd GirderBIDS

Then, you can follow the plugin instructions to run your Girder backend.

Then you can fill in the configuration file:

backend:
    type: girder
    api_url: the api URL of your girder. For instance : http://localhost:8080/api/v1
    api_key: the API key of a girder user (optional, for dev purposes)

Run the EEGAnnotator

girdereegannotator

To test the application, you can try to use the example.neonatal file provided in the assets. You can add --server to your command line to prevent your browser from opening and --port to specify the port the server should listen to, default is 8080.

Acknowledgement

This work was supported by the Agence Nationale de la Recherche (Grant ANR-22-CE45-0034).