Optimal Deep-Learning-Enabled Intelligent Decision Support System for SARS-CoV-2 Classification.

Area of Science:

• Medical Imaging and Artificial Intelligence
• Computational Biology and Bioinformatics
• Public Health and Epidemiology

Background:

• Intelligent decision support systems (IDSS) are crucial for analyzing complex healthcare data to aid medical professionals.
• The global spread of SARS-CoV-2 (COVID-19) necessitates rapid and accurate diagnostic tools.
• Computed tomography (CT) imaging shows promise for detecting SARS-CoV-2, emphasizing the need for early identification.

Purpose of the Study:

• To develop an efficient deep-learning-based fusion model with swarm intelligence (EDLFM-SI) for the identification and classification of SARS-CoV-2 infections.
• To enhance the accuracy and efficiency of COVID-19 detection using medical imaging data.

Main Methods:

• The proposed EDLFM-SI technique integrates data augmentation, preprocessing, feature extraction, and classification.
• A fusion of capsule network (CapsNet) and MobileNet is employed for feature extraction.
• The water strider algorithm (WSA) is utilized for hyperparameter tuning of deep learning models, followed by a cascaded neural network (CNN) classifier.

Main Results:

• Simulations were conducted on COVID-19 and SARS-CoV-2 CT scan datasets to evaluate the EDLFM-SI technique.
• The proposed EDLFM-SI model demonstrated superior performance compared to existing approaches in detecting SARS-CoV-2.

Conclusions:

• The EDLFM-SI technique provides an effective deep-learning solution for SARS-CoV-2 detection using CT imaging.
• The study highlights the potential of AI-driven decision support systems in combating infectious diseases like COVID-19.