dr-joint-learning

Official Repository for DRG-Net

Tusfiqur, Hasan Md, Duy MH Nguyen, Mai TN Truong, Triet A. Nguyen, Binh T. Nguyen, Michael Barz, Hans-Juergen Profitlich et al. "DRG-Net: interactive joint learning of multi-lesion segmentation and classification for diabetic retinopathy grading." arXiv preprint arXiv:2212.14615 (2022). https://arxiv.org/abs/2212.14615

• Contains the training sciprts for classification and segmentation models for FGADR and EyPACS datasets
• Scripts for Additional models and other datasets will be added soon.

Installation

Recommended environment:

• python 3.8+
• pytorch 1.7.1+
• torchvision 0.8.2+
• tqdm
• munch
• packaging
• tensorboard
• scikit-learn
• opencv-python
• pillow < 7
• segmentation-models-pytorch

To install the dependencies create a virtualenv and run:

$ pip install -r requirements.txt

Datasets

Datasets used in this paper can be accessed from the following links:

• Indian Diabetic Retinopathy Image Dataset (IDRiD) https://ieee-dataport.org/open-access/indian-diabetic-retinopathy-image-dataset-idrid
• FGADR Dataset - Look Deeper into Eyes. https://csyizhou.github.io/FGADR/
• EyePACS Dataset - https://paperswithcode.com/dataset/kaggle-eyepacs

For classification

• 'dr_classification' folder contains scripts for dr grading classification task.
• Follow readme.md inside 'dr_classification' for instructions.

For segmentation

• 'dr_segmentation' folder contains scripts for dr segmentation task.
• Follow readme.md inside 'dr_segmentaion' instructions

Citation

If you use this code or results in your research, please cite:

@article{tusfiqur2022drg,
  title={DRG-Net: interactive joint learning of multi-lesion segmentation and classification for diabetic retinopathy grading},
  author={Tusfiqur, Hasan Md and Nguyen, Duy MH and Truong, Mai TN and Nguyen, Triet A and Nguyen, Binh T and Barz, Michael and Profitlich, Hans-Juergen and Than, Ngoc TT and Le, Ngan and Xie, Pengtao and others},
  journal={arXiv preprint arXiv:2212.14615},
  year={2022}
}