[NeurIPS 2024] Samba: Severity-aware Recurrent Modeling for Cross-domain Medical Image Grading

This is the official implementation of our work entitled as Samba: Severity-aware Recurrent Modeling for Cross-domain Medical Image Grading, which has been accepted by NeurIPS 2024.

Environment Configuration

Please refer to the requirements.txt file in this project.

Training

An command example to train the model when using APTOS as the source domain is:

python main_fundus.py --cfg configs/vssm1/vssm_base_224.yaml --resume pretrain/vssm_base_0229_ckpt_epoch_237.pth --batch-size 16 --data-path G:/NeruIPS2024/datasets/DGDR/reimages/APTOS --output outputs/vssm_base_APTOS_em --dataset fundus

Please remember to change the file folder to your own.

Inference

An command example to infer the model when using DEEPDR as the unseen target domain is:

python main_fundus.py --cfg configs/vssm1/vssm_base_224.yaml --resume outputs_fundus_ema/vssm_base_DEEPDR_em/vssm1_base_0229/20240329143126/ckpt_epoch_299.pth --batch-size 16 --data-path G:/NeruIPS2024/datasets/DGDR/sdg_reimages/no_DEEPDR --output results/vssm_base_DEEPDR_em --eval --dataset fundus

Please remember to change the file folder to your own.

Acknowledgement

The development of Samba largely relies on the source code from VMamba, with the code link [https://github.com/MzeroMiko/VMamba]. We sincerely appreciate the authors for advancing Selective State Space Model (Mamba) to the visual domain.

Cite the proposed Samba

If you find the proposed Samba is useful for your task please cite our work as follows:

@inproceedings{bi2024Samba,
  title={Samba: Severity-aware Recurrent Modeling for Cross-domain Medical Image Grading},
  author={Bi, Qi and Yi, Jingjun and Zheng, Hao and Ji, Wei and Zhan, Haolan and Huang, Yawen and Li, Yuexiang and Zheng, Yefeng},
  booktitle={Advances in Neural Information Processing Systems (NeurIPS)},
  volume={37},
  year={2024}
}