Covid-19 Diagnosis by WE-SAJ

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computational Biology

Background:

• The COVID-19 pandemic presents a critical challenge due to rapidly increasing patient numbers and strained medical resources.
• Fast and accurate diagnosis of COVID-19 is essential for effective patient management and resource allocation.
• Artificial intelligence (AI) offers potential for rapid and accurate classification of medical images, including CT scans for COVID-19 detection.

Purpose of the Study:

• To propose a novel deep learning model, WE-SAJ, for the accurate classification of COVID-19 from CT images.
• To evaluate the performance of the proposed WE-SAJ model against a Jaya-based model for medical image classification.
• To demonstrate the effectiveness of the adaptive Jaya algorithm in training AI models for COVID-19 diagnosis.

Main Methods:

• Development of a deep learning model (WE-SAJ) incorporating wavelet entropy for feature extraction.
• Utilization of two-layer Feedforward Neural Networks (FNNs) for image classification.
• Application of the adaptive Jaya algorithm for training the deep learning model.

Main Results:

• The WE-SAJ model achieved a sensitivity of 85.47±1.84, specificity of 87.23±1.67, and accuracy of 86.35±0.70.
• The model demonstrated superior performance compared to the Jaya-based model in COVID-19 classification.
• Key performance metrics include precision (87.03±1.34), F1 score (86.23±0.77), and Matthews correlation coefficient (72.75±1.38).

Conclusions:

• The proposed WE-SAJ deep learning model shows significant potential for AI-driven COVID-19 diagnosis using CT images.
• The adaptive Jaya algorithm proves effective for training AI models in medical image classification tasks, outperforming the standard Jaya algorithm.
• AI techniques, particularly the developed WE-SAJ model, can aid in the rapid and accurate diagnosis of COVID-19, supporting healthcare systems.