Automated Deep Transfer Learning-Based Approach for Detection of COVID-19 Infection in Chest X-rays

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Infectious Disease Diagnostics

Background:

• Real-time polymerase chain reaction (RT-PCR) is the standard for COVID-19 detection but is slow, costly, and prone to false negatives.
• Radiological imaging, particularly chest X-rays, offers an alternative for COVID-19 diagnosis due to accessibility and lower radiation compared to CT scans.
• Manual analysis of chest X-rays for COVID-19 signatures is time-consuming and error-prone, necessitating automated solutions.

Purpose of the Study:

• To develop an automated deep transfer learning approach for detecting COVID-19 infection using chest X-ray images.
• To leverage the Xception model for enhanced COVID-19 detection accuracy and efficiency.
• To provide a faster and more reliable diagnostic tool compared to traditional methods.

Main Methods:

• Utilized chest X-ray images as the primary data source for COVID-19 detection.
• Implemented a deep transfer learning strategy employing the Xception model.
• Trained and fine-tuned the Xception network on relevant datasets for optimal performance.

Main Results:

• The proposed automated Xception model demonstrated significantly superior performance in COVID-19 detection from chest X-rays.
• Achieved higher accuracy and potentially reduced analysis time compared to existing diagnostic models.
• Validated the effectiveness of deep transfer learning for analyzing radiological signatures of COVID-19.

Conclusions:

• Automated analysis of chest X-rays using deep transfer learning, specifically the Xception model, is a viable and effective method for COVID-19 detection.
• This approach offers a promising alternative to RT-PCR, improving diagnostic speed and accuracy.
• Further research into deep learning for medical image analysis can enhance infectious disease diagnostics.