Abstract:The COVID-19 worldwide pandemic has become a great challenge for medical systems. Early detection of coronavirus increases the cure rate and saves the lives of patients. Therefore, a computer-assisted diagnostic tool is necessary to assist radiologists to quickly classify cases of pneumonia as COVID-19 or not. In this paper, we developed an approach that automatically detects COVID-19 disease. This approach relies on an optimized convolutional neural networks (CNNs) to extract deep features from both the computed tomography (CT) scans and the X-ray images. Then, features selection process is applied on CT and X-ray features to eliminate the redundant and irrelevant features and also select more accurate, appropriate, and discriminative ones. The selected CT and X-ray features are combined together using features fusion method to form a final deep feature descriptor for classification. CT and X-ray features are extracted using an optimized CNN architecture model containing thirteen convolutional layers. Sparse Filtering (SF), Features Transformation (FT), Particle Swarm Optimization (PSO), Grey Wolf Optimization (GWO), Hybrid Whale Optimization (HWO), and Entropy Mutual Information (EMI) Techniques are utilized for features selection. Furthermore, various features fusion algorithms are evaluated such as; CNN-based Fusion, Fuzzy Fusion, Canonical Correlation, Neuro Fuzzy, Curvelet based Fusion, and Discrete Wavelet Transform (DWT). Besides, for both augmented and un-augmented CT and X-ray images, we evaluate various optimization algorithms (OA), mini-batch size (M-B) values, and learning rates (LR) to reveal the optimal parameters of the proposed CNN architecture model. Regarding optimization, adaptive moment estimation (Adam) algorithm showed better performance than root mean square propagation (RMS prop), and stochastic gradient descent with momentum (SGDM). The proposed approach achieves a remarkable accuracy (99.43%) for augmented images using Hybrid Whale Optimization (HWO) technique with Canonical Correlation based fusion approach. The proposed approach is superior to traditional pre-trained CNNs and the state-of-the-art classification techniques.

An efficient hybrid classification approach for COVID-19 based on Harris Hawks Optimization and Salp Swarm Optimization

Hybrid Whale Optimization and Canonical Correlation based COVID-19 Classification Approach

A hybrid Harris Hawks optimization algorithm with simulated annealing for feature selection

Breast cancer segmentation using hybrid HHO-CS SVM optimization techniques

An Efficient Hybrid Sine-Cosine Harris Hawks Optimization For Low And High-Dimensional Feature Selection

Boosted Binary Harris Hawks Optimizer and Feature Selection

C-HMOSHSSA: Gene selection for cancer classification using multi-objective meta-heuristic and machine learning methods

Enhancing classification of lung diseases by optimizing training hyperparameters of the deep learning network

Nature-Inspired Algorithms-Based Optimal Features Selection Strategy for COVID-19 Detection Using Medical Images

Optimal feature selection for COVID-19 detection with CT images enabled by metaheuristic optimization and artificial intelligence

Hybrid optimal feature selection-based iterative deep convolution learning for COVID-19 classification system

A self-adaptive Harris Hawks optimization algorithm with opposition-based learning and chaotic local search strategy for global optimization and feature selection

Multi-strategy augmented Harris hawks optimization: performance design for feature selection

A Multi-Leader Harris Hawk Optimization Based on Differential Evolution for Feature Selection and Prediction Influenza Viruses H1N1.

Novel Crow Swarm Optimization Algorithm and Selection Approach for Optimal Deep Learning COVID-19 Diagnostic Model

A Hybrid COVID-19 Detection Model Using an Improved Marine Predators Algorithm and a Ranking-Based Diversity Reduction Strategy

HMOSHSSA: a novel framework for solving simultaneous clustering and feature selection problems

Automatic Classification of Covid-19 Chest X-Ray Images Using Local Binary Pattern and Binary Particle Swarm Optimization for Feature Selection

An Automated Glowworm Swarm Optimization with an Inception-Based Deep Convolutional Neural Network for COVID-19 Diagnosis and Classification

Lung infection detection and classification using the integration of the improved grasshopper and the remora optimization approaches with improved SVM

A comparative analysis of meta-heuristic optimization algorithms for feature selection on ML-based classification of heart-related diseases