Computer Aided COVID-19 Diagnosis in Pandemic Era Using CNN in Chest X-ray Images.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Radiology

Background:

• Early detection of COVID-19 via chest X-rays is crucial for pandemic management and patient treatment.
• Accurate analysis of chest X-rays aids in contact tracing and effective disease control strategies.

Purpose of the Study:

• To present a computationally efficient deep hybrid learning framework for COVID-19 detection using chest X-ray images.
• To develop a novel Convolutional Neural Network (CNN) named COV-Net for identifying COVID-specific patterns.

Main Methods:

• A novel COV-Net architecture was developed for analyzing chest X-ray images, incorporating max-pooling for pattern boundary learning.
• The framework integrates COV-Net with machine learning classifiers, including support vector machines, for enhanced discrimination.
• The model was evaluated on a public dataset comprising X-rays of COVID-19, pneumonia, and healthy patients.

Main Results:

• The proposed deep hybrid learning method achieved high performance metrics: 96.69% recall, 96.72% precision, 96.73% accuracy, and 96.71% F-score for COVID-19 detection.
• For multi-class and binary classification (COVID-19 vs. pneumonia), the model demonstrated superior results with 99.21% recall, 99.22% precision, 99.21% F-score, and 99.23% accuracy.

Conclusions:

• The developed deep hybrid learning framework, featuring the COV-Net, offers a computationally light and effective solution for COVID-19 detection from chest X-rays.
• The framework's high accuracy in classifying COVID-19, pneumonia, and normal cases supports its utility in clinical settings for early diagnosis and management.