An Efficient Deep Learning Model to Detect COVID-19 Using Chest X-ray Images

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Infectious Diseases

Background:

• The COVID-19 pandemic caused global healthcare disruptions.
• Current COVID-19 testing methods face challenges and controversies.
• There is a need for improved, cost-effective disease detection.

Purpose of the Study:

• To develop and evaluate a Deep Learning Method (DLM) for COVID-19 detection using chest X-ray (CXR) images.
• To assess the efficacy of machine learning (ML) as a forecasting tool for COVID-19 diagnosis.
• To provide a rapid and accessible method for identifying potential COVID-19 cases.

Main Methods:

• Utilized a dataset of 10,040 chest X-ray images, including COVID-19 positive (2143), pneumonia (3674), and normal (4223) cases.
• Applied a Deep Learning Method (DLM) for image analysis and classification.
• Evaluated model performance using accuracy, sensitivity, and ROC curve analysis.

Main Results:

• The DLM achieved a COVID-19 detection accuracy of 96.43% and a sensitivity of 93.68%.
• The model demonstrated high performance with an Area Under the ROC Curve (AUC) of 99% for COVID-19, 97% for pneumonia, and 98% for normal cases.
• Chest X-ray images proved effective for COVID-19 detection, outperforming more expensive and time-consuming pathological tests.

Conclusions:

• Machine learning approaches offer a viable solution for rapid analysis of CXR images.
• This DLM can assist radiologists in efficiently filtering potential COVID-19 candidates.
• The study highlights the potential of AI in enhancing diagnostic capabilities during pandemics.