1. Open terminal and setup 3DUNetCNN repository:

git clone https://github.com/ellisdg/3DUnetCNN.git
cd 3DUnetCNN
export PYTHONPATH=${PWD}:${PYTHONPATH}

2. cd into the brats2020 example directory:

cd examples/brats2020

3. Download the BraTS2020 data to the brats2020 example directory. The training data folder should be named MICCAI_BraTS2020_TrainingData and the validation data folder should be named MICCAI_BraTS2020_ValidationData.

4. (optional) Setup the configuration file using the "create_config_lomem.ipynb" notebook See the "create_config_lomem.ipynb" notebook for details on creating the configuration file. If the data is in the brats2020 folder and that is the current directory, then this step should be optional. However, if the data is somewhere else, or the training script cannot find the data, then modify that notebook to point to the correct location for the training/validation data.

5. Run the training

python /path/to/unet3d/scripts/train.py --config_filename brats2020_config.json

You can also set the number of gpus '--ngpus' and the number of threads '--nthreads' to use during training. The outputs will be written to a folder called 'brats2020_config' based on the name and location of the configuration file. If cross-validation is used, then there will be separate folders within that folder where the model, training, and inference results will be saved. Any filenames in training script will also try to predict the outputs of any of keys listed in the configuration as "_filenames" except for the "training_filenames". For the configuration given, this means that there will be prediction folders for bratsvalidation and validation sets.