Related Experiment Video
Updated: Oct 31, 2025
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.4K
Transfer Learning to Detect COVID-19 Automatically from X-Ray Images Using Convolutional Neural Networks
Mundher Mohammed Taresh1, Ningbo Zhu1, Talal Ahmed Ali Ali1
1College of Information Science and Engineering, Hunan University, Changsha 400013, China.
International Journal of Biomedical Imaging
|July 1, 2021
Summary
Artificial intelligence using deep learning on chest X-rays accurately detects COVID-19. VGG16 and MobileNet models show high performance, aiding rapid diagnosis and improving patient outcomes.
Area of Science:
- Medical Imaging and Artificial Intelligence
- Infectious Disease Diagnostics
Background:
- The COVID-19 pandemic necessitates rapid and accurate diagnostic tools.
- Chest X-rays (CXRs) are a common imaging modality for respiratory illnesses.
- Deep learning, particularly Convolutional Neural Networks (CNNs), offers potential for automated image analysis.
Purpose of the Study:
- To evaluate the effectiveness of pretrained CNNs for automated COVID-19 diagnosis from CXRs.
- To identify the best-performing deep learning algorithms for COVID-19 detection using CXR images.
- To assess the utility of AI in identifying biological markers of COVID-19 from X-rays.
Main Methods:
- Utilized a dataset comprising 1200 COVID-19, 1345 viral pneumonia, and 1341 healthy CXR images.
- Applied and fine-tuned state-of-the-art pretrained CNN models, including VGG16 and MobileNet.
- Evaluated model performance based on accuracy, F1 score, precision, specificity, and sensitivity.
Main Results:
- Deep learning models demonstrated significant utility in identifying COVID-19 markers from CXRs.
- VGG16 and MobileNet achieved high accuracy rates of 98.28% during initial evaluation.
- VGG16 exhibited superior performance with an accuracy of 98.72%, F1 score of 97.59%, precision of 96.43%, specificity of 98.70%, and sensitivity of 98.78%.
Conclusions:
- Pretrained deep learning models, especially VGG16, can significantly enhance the speed and accuracy of COVID-19 diagnosis from CXRs.
- AI-powered analysis of X-ray images holds promise for clinical decision support in managing the pandemic.
- Larger datasets are crucial for further improving the reliability of deep transfer learning models for COVID-19 identification.
