MedShift

The official implementation of MedShift: Implicit Conditional Transport for X-Ray Domain Adaptation.

Francisco Caetano¹, Christiaan Viviers¹, Peter H.N. de With¹, Fons van der Sommen¹

¹ Eindhoven University of Technology

Prerequisites

You will need:

• python (see pyproject.toml for full version)
• Git
• Make
• a .secrets file with the required secrets and credentials
• load environment variables from .env
• NVIDIA Drivers(mandatory) and CUDA >= 12.6 (mandatory if Docker/Apptainer is not used)
• Weights & Biases account

Installation

Clone this repository (requires git ssh keys)

git clone --recursive git@github.com:caetas/MedShift.git
cd MedShift

Using Docker or Apptainer

Create a .secrets file and add your Weights & Biases API Key:

WANDB_API_KEY = <your-wandb-api-key>

Docker

Create the image using the provided Dockerfile

docker build --tag medshift .

Or download it from the Hub:

docker pull docker://ocaetas/medshift

Then run the script job_docker.sh that will execute main.sh:

cd scripts
bash job_docker.sh

To access the shell, please run:

docker run --rm -it --gpus all --ipc=host --env-file .env -v $(pwd)/:/app/ medshift bash

Apptainer

Convert the Docker Image to a .sif file:

apptainer pull medshift.sif docker://ocaetas/medshift

Then run the script job_apptainer.sh that will execute main.sh:

cd scripts
bash job_apptainer.sh

To access the shell, please run:

apptainer shell --nv --env-file .env --bind $(pwd)/:/app/ medshift.sif

Add the flag --nvccli if you are using WSL.

Note: Edit the main.sh script if you want to train a different model.

Normal Installation

Create the Conda Environment:

conda env create -f environment.yml
conda activate python3.11

On Linux

And then setup all virtualenv using make file recipe

(python3.11) $ make setup-all

You might be required to run the following command once to setup the automatic activation of the conda environment and the virtualenv:

direnv allow

Feel free to edit the .envrc file if you prefer to activate the environments manually.

On Windows

You can setup the virtualenv by running the following commands:

python -m venv .venv-dev
.venv-dev/Scripts/Activate.ps1
python -m pip install --upgrade pip setuptools
python -m pip install -r requirements/requirements.txt

To run the code please remember to always activate both environments:

conda activate python3.11
.venv-dev/Scripts/Activate.ps1

Dataset

The dataset should be downloaded here. Unzip the dataset and move the folders to data/raw.

Training the Models

In addition to the instructions for using Docker or Apptainer, the documentation for training is available here: TRAINING.md.

Running and Evaluating the Models

The pretrained model can be downloaded here.

The instructions to run and evaluate the models are available in INFERENCE.md.

License

This project is licensed under the terms of the MIT license. See LICENSE for more details.

Citation

If you publish work that uses SymmFlow, please cite SymmFlow as follows:

Will be added later.