NKI-AI/direct-cmrxrecon-challenge23

DIRECT — CMRxRecon Challenge 2023 (vSHARP)

vSHARP checkpoint released for the CMRxRecon Challenge 2023.

• Method: vSHARP
• Challenge: cmrxrecon.github.io
• Previous zip: files.aiforoncology.nl/.../vsharp_2D_dynamic.zip
• Docs: DIRECT model zoo

Dataset

Dataset	CMRxRecon 2023 challenge data
Domain	Multi-coil cardiac MRI (cine / mapping tasks)
Site	cmrxrecon.github.io

Request access and download instructions are on the challenge website. Point --data-root at your local challenge layout.

Files

vsharp_2d_dynamic.yaml
vsharp_2d_dynamic.pt

Masking

Challenge undersampling masks ship with the data (YAML extra_keys, e.g. mask04). Do not add a multi-acceleration training masking.accelerations list for inference — use the released challenge mask keys.

Install DIRECT

git clone https://github.com/NKI-AI/direct.git
cd direct
conda create --name direct python=3.12
conda activate direct
pip install meson-python meson ninja
pip install --no-build-isolation -e ".[dev]"

Usage

hf download NKI-AI/direct-cmrxrecon-challenge23 --local-dir ./cmrx23

direct predict ./predictions \
  --cfg ./cmrx23/vsharp_2d_dynamic.yaml \
  --checkpoint ./cmrx23/vsharp_2d_dynamic.pt \
  --data-root /path/to/cmrxrecon \
  --num-gpus 1

The first argument is the prediction output directory.

Citation

If you use these models or DIRECT, please cite the DIRECT toolkit and the method paper(s) below.

DIRECT

@article{DIRECTTOOLKIT,
  title={DIRECT: Deep Image REConstruction Toolkit},
  author={Yiasemis, George and Moriakov, Nikita and Karkalousos, Dimitrios and Caan, Matthan and Teuwen, Jonas},
  journal={Journal of Open Source Software},
  volume={7},
  number={73},
  pages={4278},
  year={2022},
  doi={10.21105/joss.04278},
  url={https://doi.org/10.21105/joss.04278}
}

Method

@inproceedings{yiasemis2023deep,
  title={Deep cardiac {MRI} reconstruction with {ADMM}},
  author={Yiasemis, George and Moriakov, Nikita and Sonke, Jan-Jakob and Teuwen, Jonas},
  booktitle={International Workshop on Statistical Atlases and Computational Models of the Heart},
  pages={479--490},
  year={2023},
  organization={Springer}
}

@article{yiasemis2024vsharp,
  title={vSHARP: variable Splitting Half-quadratic {ADMM} algorithm for Reconstruction of inverse Problems},
  author={Yiasemis, George and Moriakov, Nikita and S{\'a}nchez, Clara I. and Sonke, Jan-Jakob and Teuwen, Jonas},
  journal={Magnetic Resonance Materials in Physics, Biology and Medicine},
  year={2024},
  doi={10.1007/s10334-024-01189-0}
}