A genetic programming-based convolutional deep learning algorithm for identifying COVID-19 cases via X-ray images.

Area of Science:

• Artificial Intelligence
• Medical Imaging
• Computational Biology

Background:

• Convolutional Neural Networks (CNNs) are effective for image processing tasks.
• Optimizing CNN architecture is critical for accuracy and computational efficiency.
• COVID-19 diagnosis from X-ray images requires robust and efficient detection methods.

Purpose of the Study:

• To develop a genetic programming approach for optimizing CNN structures.
• To enhance the accuracy and reduce the computational cost of COVID-19 diagnosis using X-ray images.
• To propose a novel method for automatic CNN architecture design.

Main Methods:

• A genetic programming framework was developed for CNN structure optimization.
• A graph representation for CNN architecture was proposed.
• Evolutionary operators (crossover, mutation) were designed for the graph representation.
• A co-evolutionary scheme optimized both CNN skeleton and numerical parameters.

Main Results:

• The proposed method successfully optimized CNN architectures for COVID-19 detection.
• The approach demonstrated potential for improving diagnostic accuracy.
• The evolutionary optimization addressed both structural arrangement and operator properties.

Conclusions:

• Genetic programming offers a flexible and effective approach for optimizing CNN architectures.
• The proposed method is a promising tool for medical image analysis, specifically for COVID-19 diagnosis.
• Automated CNN architecture optimization can lead to more efficient and accurate diagnostic tools.