All the codes are in the code/ folder.
We use the KiTS19 dataset and preprocessing code provided by nnUNet repository.
-
download the KiTS19 dataset from the above repository. The raw data are in the data/ folder.
-
Organized the raw data according to the requirements of nnUNet.
We created a script for this. Just run python3 reorganize_data_folder.py. nnUNet also requires a json file for data preprocess, run create_json.py to generate it. You have to change the path of folders where you want to store the datas accordingly. The lines need to be changed have been marked by comments.
- Preprocessing
First, you need to install the nnUNet according to the instruction of this repo. After doing step 2, run command line nnUNet_convert_decathlon_task -i FOLDER_TO_TASK_AS_DOWNLOADED_FROM_MSD -p NUM_PROCESSES and nnUNet_plan_and_preprocess -t XXX --verify_dataset_integrity to get the preprocessed data. Detailed descriptions can be found here.
After the data are ready, run python3 train.py [task_name] to train and test the model. [task_name] is the identifier for your training settings. The settings can be changed in train.py file, where marked by comments.