code for the paper 'Few-shot Breast Cancer Metastases Classification via Unsupervised Cell Ranking'

our method are implemented in the pytorch framework,and the data processing are mainly referred to here

pre-processing

extract patches for training

post-processing

generate heatmap and extract features from heatmap

ranking.py

train the ranking model

classification.py

train the supervised model

finetune.py

finetune the pre-train ranking model

References

@article{title={Few-shot Breast Cancer Metastases Classification via Unsupervised Cell Ranking},
  author={Jiaojiao Chen, Jianbo Jiao, Shengfeng He, Guoqiang Han, and Jing Qin},
  journal={IEEE/ACM Transactions on Computational Biology and Bioinformatics},
  year={2019}
}