Pytorch implementation of CANet: Cross-disease attention network.

CANet: Cross-disease Attention Network for Joint Diabetic Retinopathy and Diabetic Macular Edema Grading
IEEE Transactions on Medical Imaging, 2019

• Install Pytorch 1.1.0 and CUDA 9.0
• Clone this repo

git clone https://github.com/xmengli999/CANet
cd CANet

• Download Messidor dataset
  
• Put the data under ./data/

• Download ImageNet pretrain model and put it under ./pretrain/ or Download the kaggle DR pretrain model and put it under ./pretrain/
• cd messidor_scripts and specify the pretrain model in train_fold.sh
• Run sh train_fold.sh to start the training process

• Specify the model path in eval_fold.sh
• Run sh eval_fold.sh to start the evaluation.

If you find the code useful for your research, please cite our paper.

@article{li2019canet,
  title={CANet: Cross-Disease Attention Network for Joint Diabetic Retinopathy and Diabetic Macular Edema Grading},
  author={Li, Xiaomeng and Hu, Xiaowei and Yu, Lequan and Zhu, Lei and Fu, Chi-Wing and Heng, Pheng-Ann},
  journal={IEEE transactions on medical imaging},
  volume={39},
  number={5},
  pages={1483--1493},
  year={2019},
  publisher={IEEE}
}

CBAM module is reused from the Pytorch implementation of CBAM.

• Contact: Xiaomeng Li ([email protected])