An effective COVID-19 classification in X-ray images using a new deep learning framework.

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computer-Aided Diagnosis

Background:

• Growing global concern over lung-related diseases, exacerbated by COVID-19 transmission.
• Artificial Intelligence (AI) offers rapid identification of COVID-19 from chest X-rays.

Purpose of the Study:

• To develop and evaluate a hybrid deep-learning approach for diagnosing lung disorders using chest X-ray images.
• To enhance the accuracy and efficiency of disease detection through advanced AI techniques.

Main Methods:

• Utilized a public COVID-19 Chest X-ray dataset.
• Pre-processed images with Improved Anisotropic Diffusion Filtering (IADF).
• Employed feature extraction methods: GLCM, uLBP, HoG, hvnLBP.
• Optimized feature selection using Adaptive Reptile Search Optimization (ARSO).
• Developed a Multi-head Attention-based Bi-directional Gated Recurrent unit with Deep Sparse Auto-encoder Network (MhA-Bi-GRU with DSAN) for classification.
• Applied Dynamic Levy-Flight Chimp Optimization (DLF-CO) to minimize the loss function.

Main Results:

• Achieved high classification accuracy: 0.95% at a 0.001 learning rate and 0.98% at a 0.0001 learning rate.
• The proposed methodology demonstrated superior performance compared to existing methods across various parameters.
• Effective diagnosis of lung diseases using chest X-ray images was confirmed.

Conclusions:

• The proposed hybrid deep-learning approach, integrating advanced feature extraction and optimal selection, effectively diagnoses lung diseases from X-ray images.
• This AI-driven method shows significant potential for improving the accuracy and speed of medical diagnoses.