Related Experiment Video
Updated: Aug 5, 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
14.2K
Design and Analysis of a Deep Learning Ensemble Framework Model for the Detection of COVID-19 and Pneumonia Using
Xingsi Xue1, Seelammal Chinnaperumal2, Ghaida Muttashar Abdulsahib3
1Fujian Provincial Key Laboratory of Big Data Mining and Applications, Fujian University of Technology, Fuzhou 350011, China.
Bioengineering (Basel, Switzerland)
|March 29, 2023
Summary
Deep learning models accurately detect COVID-19 and pneumonia using chest X-rays and CT scans. Enhanced VGG16 achieved 99% accuracy, outperforming other methods for viral infection diagnosis.
Area of Science:
- Medical Imaging Analysis
- Artificial Intelligence in Healthcare
- Deep Learning Applications
Background:
- Accurate COVID-19 detection is crucial, with PCR testing and medical imaging like X-rays and CT scans being primary methods.
- Deep learning (DL) and artificial intelligence (AI) offer powerful tools for early and precise identification of COVID-19 and related conditions.
- Existing DL models require optimization for medical image analysis tasks.
Purpose of the Study:
- To explore and evaluate various deep learning techniques for identifying COVID-19 and pneumonia from medical CT and radiography images.
- To automate the selection of optimal model architectures and training parameters for diagnostic models.
- To leverage transfer learning to address data limitations and reduce model training time.
Main Methods:
- Utilized deep learning models including ResNet152, VGG16, ResNet50, and DenseNet121 for image analysis.
- Applied transfer learning approaches, including an enhanced VGG16 architecture, for multi-class classification of radiographic images.
- Validated model performance using publicly available X-ray and CT scan datasets, assessing metrics like accuracy and F-score.
Main Results:
- The enhanced VGG16 model demonstrated high accuracy (99%) in recognizing typical radiographic images for COVID-19 and pneumonia.
- The proposed models, particularly the ResNet framework for CT scans, showed strong performance in accuracy and precision.
- Achieved an average F-score of 95% and 97%, indicating superior diagnostic capability compared to competing methods.
Conclusions:
- The developed deep learning models are highly effective for the recognition and classification of COVID-19 and pneumonia from medical images.
- The enhanced VGG16 architecture provides an efficient and accurate solution for multi-class classification of various radiographic findings.
- This research offers a more efficient methodology for coronavirus detection compared to existing approaches, especially for viral infections.
