Project for ML4Med in WS 2023
If you encounter any errors, problems or have suggestions, please feel free to improve this read me.
look for "data_dir = r'C:\Users\Nicolai\PycharmProjects\ML4MedWS2023\Data\ConvertedPelvisNiftiDataset\Converted Pelvis Nifti Dataset'"
if the vanilla training runs, try out your self different parameters and models and so on. For that please copy "Code/MONAI/your_custom_training.py" and rename it and then customize it.
https://universe.roboflow.com/ks-fsm9o/pelvis-ap-x-ray/dataset/2
https://cloud.rz.uni-kiel.de/index.php/s/WRM33znRQjLcMZo
the data ist not included in this repository, because it is too large. but you can automatically download it by running "Code/GetData/GetData.py" in "Code/GetData"
https://github.com/Project-MONAI/tutorials/blob/main/3d_segmentation/spleen_segmentation_3d.ipynb
- Open Project in PyCharm
- Open Settings -> Project -> Python Interpreter
- Add new Python Interpreter
- Select "Add Local Interpreter"
- Select "New environment using Virtualenv"
- Select a Python Version (e.g. 3.11)
- Now you have a clean new Python Environment for a specific Python version
- Install all required packages via 'pip install -q "monai-weekly[gdown, nibabel, tqdm, ignite]"' and 'pip install -q matplotlib SimpleITK pandas scipy'
docker build -t ml4med:latest .
docker run --gpus device=1 -v /home/sukin699/code/ML4MedWS2023/Data/ConvertedPelvisNiftiDataset:/code/ML4MedWS2023/Data/ConvertedPelvisNiftiDataset -i -t ml4med:latest bash