COVID-Nets: deep CNN architectures for detecting COVID-19 using chest CT scans

Hammam Alshazly; Christoph Linse; Mohamed Abdalla; Erhardt Barth; Thomas Martinetz

doi:10.7717/peerj-cs.655

COVID-Nets: deep CNN architectures for detecting COVID-19 using chest CT scans

PeerJ Comput Sci. 2021 Jul 29:7:e655. doi: 10.7717/peerj-cs.655. eCollection 2021.

Authors

Hammam Alshazly^{1

2}, Christoph Linse¹, Mohamed Abdalla^{3

4}, Erhardt Barth¹, Thomas Martinetz¹

Affiliations

¹ Institut für Neuro- und Bioinformatik, University of Lübeck, Lübeck, Germany.
² Faculty of Computers and Information, South Valley University, Qena, Egypt.
³ Mathematics Department, Faculty of Science, King Khalid University, Abha, Saudi Arabia.
⁴ Mathematics Department, Faculty of Science, South Valley University, Qena, Egypt.

Abstract

In this paper we propose two novel deep convolutional network architectures, CovidResNet and CovidDenseNet, to diagnose COVID-19 based on CT images. The models enable transfer learning between different architectures, which might significantly boost the diagnostic performance. Whereas novel architectures usually suffer from the lack of pretrained weights, our proposed models can be partly initialized with larger baseline models like ResNet50 and DenseNet121, which is attractive because of the abundance of public repositories. The architectures are utilized in a first experimental study on the SARS-CoV-2 CT-scan dataset, which contains 4173 CT images for 210 subjects structured in a subject-wise manner into three different classes. The models differentiate between COVID-19, non-COVID-19 viral pneumonia, and healthy samples. We also investigate their performance under three binary classification scenarios where we distinguish COVID-19 from healthy, COVID-19 from non-COVID-19 viral pneumonia, and non-COVID-19 from healthy, respectively. Our proposed models achieve up to 93.87% accuracy, 99.13% precision, 92.49% sensitivity, 97.73% specificity, 95.70% F1-score, and 96.80% AUC score for binary classification, and up to 83.89% accuracy, 80.36% precision, 82.04% sensitivity, 92.07% specificity, 81.05% F1-score, and 94.20% AUC score for the three-class classification tasks. We also validated our models on the COVID19-CT dataset to differentiate COVID-19 and other non-COVID-19 viral infections, and our CovidDenseNet model achieved the best performance with 81.77% accuracy, 79.05% precision, 84.69% sensitivity, 79.05% specificity, 81.77% F1-score, and 87.50% AUC score. The experimental results reveal the effectiveness of the proposed networks in automated COVID-19 detection where they outperform standard models on the considered datasets while being more efficient.

Keywords: Automated diagnosis; COVID-19 detection; Computed tomography; Deep learning; Multi-class classification; SARS-CoV-2.

Grants and funding

Hammam Alshazly and Mohamed Abdalla received funding through the Research Group Program under grant number RGP.1/3/42 from the Deanship of Scientific Research at King Khalid University. The work of Christoph Linse was supported by the Bundesministeriums für Wirtschaft und Energie (BMWi) through the Mittelstand 4.0-Kompetenzzentrum Kiel Project. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.