belal-bh / covid19-resnet50-tl

Transfer learning with fine-tuned deep CNN ResNet50 model for classifying COVID-19 from chest X-ray images

COVID19-ResNet50-TL

This repository contains code implementations and description of the research experiments for detecting COVID-19 from chest X-ray images. The research paper Transfer learning with fine-tuned deep CNN ResNet50 model for classifying COVID-19 from chest X-ray images is publicly available in the Informatics in Medicine Unlocked (Elsevier) journal.

Note: This repository will be update soon with more information. Stay tuned! Thanks

Colaboratory Notebooks

We conduct our research experiment on the Google Colaboratory runtime environment. We utilize the free GPU runtime of google colaboratory. Because of the daily free GPU limit, we conduct our experiments in different colaboratory notebooks. We create ten different notebooks for ten different Transfer Learning (TL) models. Most of the code same except for the checkpoint sections of the corresponding TL model.

Please find the details about our research colaboratory notebooks in the colab_notebooks directory of this repository or click here (colab_notebooks).

Welcome

If this research experiment helps in any of your research, please cite this paper.

Thank you.

Cite:

@article{HOSSAIN2022100916,
title = {Transfer learning with fine-tuned deep CNN ResNet50 model for classifying COVID-19 from chest X-ray images},
journal = {Informatics in Medicine Unlocked},
volume = {30},
pages = {100916},
year = {2022},
issn = {2352-9148},
doi = {https://doi.org/10.1016/j.imu.2022.100916},
url = {https://www.sciencedirect.com/science/article/pii/S235291482200065X},
author = {Md. Belal Hossain and S.M. Hasan Sazzad Iqbal and Md. Monirul Islam and Md. Nasim Akhtar and Iqbal H. Sarker},
keywords = {Deep learning, Transfer learning, ResNet50, COVID-19}
}