COVID-19 Detection by Optimizing Deep Residual Features with Improved Clustering-Based Golden Ratio Optimizer

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• The COVID-19 pandemic necessitates rapid and accurate diagnostic tools.
• Existing methods like RT-PCR, CT, and X-ray have limitations or varying effectiveness.
• Early detection is crucial due to the lack of specific antiviral treatments.

Purpose of the Study:

• To develop a cost-effective computational model for automated COVID-19 detection.
• To enhance diagnostic accuracy using deep features and a novel meta-heuristic optimization approach.
• To validate the model's performance on diverse public datasets.

Main Methods:

• Extraction of deep features from medical images (Chest X-ray and CT scans).
• Implementation of a novel meta-heuristic feature selection method: Clustering-based Golden Ratio Optimizer (CGRO).
• Validation of the proposed model on three distinct datasets: COVID CT-dataset, SARS-Cov-2 dataset, and Chest X-Ray dataset.

Main Results:

• The model achieved high detection accuracies across all tested datasets.
• Specific accuracies include 99.31% on the COVID CT-dataset, 98.65% on the SARS-Cov-2 dataset, and 99.44% on the Chest X-Ray dataset.
• The proposed CGRO approach demonstrated state-of-the-art performance in feature selection for COVID-19 detection.

Conclusions:

• The developed computational model offers a promising, less expensive solution for automated COVID-19 detection.
• The integration of deep feature extraction and CGRO significantly improves diagnostic accuracy.
• This approach provides a valuable tool for early and efficient identification of COVID-19 from radiological images.