OptCoNet

OptCoNet performs automated classification of chest X-ray images to distinguish COVID-19, pneumonia, and normal cases using an optimized convolutional neural network.

Key Features:

• Automated Diagnosis: OptCoNet distinguishes COVID-19, normal, and pneumonia cases from chest X-ray images.
• Optimized Architecture: An optimized convolutional neural network architecture enhances feature extraction and classification in medical imaging.
• Hyperparameter Optimization (Grey Wolf Optimizer): The Grey Wolf Optimizer (GWO) algorithm fine-tunes CNN hyperparameters for training the network.
• Training Dataset: The model is trained on an openly accessible chest X-ray dataset comprising COVID-19, normal, and pneumonia images.
• Performance Metrics: Reported accuracy 97.78%, sensitivity 97.75%, specificity 96.25%, precision 92.88%, and F1 score 95.25%.

Scientific Applications:

• Automated radiographic screening: Rapid classification of chest X-rays to support differential diagnosis among COVID-19, pneumonia, and normal findings.

Methodology:

The convolutional neural network is trained on an openly accessible chest X-ray dataset of COVID-19, normal, and pneumonia cases, with Grey Wolf Optimizer (GWO) used to optimize the network hyperparameters.