Deep learning model for the automatic classification of COVID-19 pneumonia, non-COVID-19 pneumonia, and the healthy: A multi-center retrospective study

This repository includes source code of our paper (Deep learning model for classification between COVID-19 pneumonia, non-COVID-19 pneumonia, and the healthy on chest X-ray image). Our model is based on EfficientNet. The detail of the model is described in our paper.

Requirement

Our source code (run_effi.py) is written in Python (version 3.7 or later). Scikit-learn, tensroflow, keras, and efficient packages are required for training and evaluating the model.

Dataset

The public datasets used for the paper are (i) COVIDx dataset and (ii) COVID_BIMCV dataset. The details of these public datasets are described in the paper.

These two public datasets are available from the following URLs.

1. COVIDx dataset (version 5): https://u.pcloud.link/publink/show?code=XZFnxnXZQkR4o66huaRvmnDzozsdgLsx7jP7

2. COVID_BIMCV dataset: https://u.pcloud.link/publink/show?code=XZznxnXZ2FjPolxpkSy7HqxPeBfUDHENfujV

Our private dataset is not disclosed because of privacy protection.

Paper

If the source code of this repository is used, please cite the following paper.

Nishio, M., Kobayashi, D., Nishioka, E. et al. Deep learning model for the automatic classification of COVID-19 pneumonia, non-COVID-19 pneumonia, and the healthy: a multi-center retrospective study. Sci Rep 12, 8214 (2022).

https://www.nature.com/articles/s41598-022-11990-3

License

The source code of this repository is licensed under GPLv3. For the detail, please see License file of this repository.