PyTorch implementation for our paper on TMI2022:
Y. Tan, K. -F. Yang, S. -X. Zhao and Y. -J. Li, "Retinal Vessel Segmentation with Skeletal Prior and Contrastive Loss," in IEEE Transactions on Medical Imaging, doi: 10.1109/TMI.2022.3161681.

WebPage:https://ieeexplore.ieee.org/abstract/document/9740153

/code

• Data Augmentation Method: code\color_space_mixture.py
• Sample Contrastive Learning: code\sample_contrastive_learning.py
• Skeletal Prior based Network: code\model_skelcon.py

/onnx

• Pytorch trained weights from DRIVE, STARE, CHASE DB1, and HRF datasets.
• The *.onnx weights can be directly used to extract vessels from fundus images, see onnx\infer.py

/results

• segmentation results for popular datasets: results\popular
• segmentation results for cross-dataset-validation: results\generalization

For any questions, please contact me. And my e-mails are

• [email protected]
• [email protected]

If you use this codes in your research, please cite the paper:

@article{tan2022retinal,
  title={Retinal Vessel Segmentation with Skeletal Prior and Contrastive Loss},
  author={Tan, Yubo and Yang, Kai-Fu and Zhao, Shi-Xuan and Li, Yong-Jie},
  journal={IEEE Transactions on Medical Imaging},
  year={2022},
  doi={10.1109/TMI.2022.3161681}
  publisher={IEEE}
}