DON: Deep Learning and Optimization-Based Framework for Detection of Novel Coronavirus Disease Using X-ray Images

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• Limited COVID-19 test kits necessitate rapid, automated diagnostic alternatives.
• Preventing the spread of coronavirus requires efficient detection methods.

Purpose of the Study:

• To develop a deep learning methodology for detecting COVID-19 infected patients using X-ray images.
• To propose a multi-objective optimization technique for enhancing diagnostic accuracy.

Main Methods:

• Utilized eleven Convolutional Neural Network (CNN) models for analyzing X-ray images.
• Employed the J48 decision tree method for classifying deep features of infected X-ray images.
• Optimized CNN parameters using the Emperor Penguin Optimizer (MOEPO).

Main Results:

• The proposed model achieved high precision, accuracy, recall, specificity, and F1-score in categorizing X-ray images.
• Demonstrated superior performance compared to existing models in extensive testing.
• The model effectively detects coronavirus pneumonia from chest X-ray images.

Conclusions:

• The developed deep learning model offers a viable solution for real-time COVID-19 detection using X-ray imaging.
• This automated method can significantly aid in controlling the spread of COVID-19.
• The multi-objective optimization enhances the reliability of AI-driven diagnostics.