Overview of deep learning models for identification Covid-19

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Infectious Disease Research

Background:

• COVID-19 pandemic significantly impacts global health and necessitates rapid diagnostic tools.
• The virus causes respiratory issues, visible as white patches on chest X-rays.
• Automatic detection systems are crucial for early identification and isolation of infected individuals.

Purpose of the Study:

• To provide an overview of modern deep learning-based methods for COVID-19 detection from X-ray images.
• To compare the diagnostic performance of various deep learning models in identifying COVID-19.
• To evaluate the effectiveness of AI in screening and mitigating the spread of COVID-19.

Main Methods:

• Utilizing deep learning (AI) techniques for analyzing chest X-ray images.
• Implementing and comparing specific deep learning models such as RestNet50 and Deep Convolutional Neural Networks (DCNN).
• Assessing diagnostic accuracy based on radiology image analysis.

Main Results:

• Deep learning models show promise for reliable and effective COVID-19 screening.
• The RestNet50 pre-trained and DCNN models achieved a diagnostic accuracy of 98%.
• This accuracy represents the highest reported performance among the discussed deep learning models.

Conclusions:

• Deep learning-based analysis of chest X-rays is a viable and highly accurate method for COVID-19 detection.
• High accuracy achieved by models like RestNet50 and DCNN can support rapid diagnosis and isolation strategies.
• AI-powered screening systems are essential tools in managing and mitigating the spread of infectious diseases like COVID-19.