COVID-19 Diagnosis from CT Images with Convolutional Neural Network Optimized by Marine Predator Optimization...

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• The global spread of Coronavirus Disease 2019 (COVID-19) necessitates rapid and accurate diagnostic tools.
• Timely diagnosis and isolation of infected individuals are critical for controlling the pandemic.
• Current diagnostic methods face challenges in speed and accessibility.

Purpose of the Study:

• To develop and evaluate a novel hybrid deep learning model for automated COVID-19 diagnosis from CT images.
• To enhance diagnostic performance through optimization using a metaheuristic algorithm.
• To compare the proposed method against existing techniques for validation.

Main Methods:

• A hybrid convolutional neural network (CNN) architecture was employed for image analysis.
• The CNN model was optimized using the marine predator optimization algorithm (MPA).
• The method was trained and validated on the MosMedData dataset of chest CT scans.

Main Results:

• The proposed hybrid CNN-MPA model achieved high performance metrics.
• Achieved accuracy of 98.11%, precision of 98.13%, sensitivity of 98.66%, and F1 score of 97.26%.
• Outperformed three other comparative methods in all evaluated indicators.

Conclusions:

• The developed AI-based method demonstrates superior accuracy and reliability for COVID-19 diagnosis using CT scans.
• This approach offers a promising tool for rapid and effective screening of COVID-19 patients.
• The integration of metaheuristic optimization significantly enhances the diagnostic capabilities of deep learning models.