Detection of COVID-19 using Chest X-Ray Scans

Abstract

COVID-19 has proven to be the unseen and unforeseen pandemic nobody was prepared to face. Healthcare professionals and radiologists have been under a lot of pressure ever since the outbreak to treat and develop faster ways to detect the disease. The proposed research involves the classification of chest X-ray scans to identify whether a patient has been infected with COVID-19 or not using the concept of Transfer Learning. Two methods of classifying the images have been presented in the research. The first approach classifies a given image into two categories being COVID-19 and Non COVID-19. The second approach classifies a given image into three categories namely COVID-19, Healthy and Pneumonia. These two models have obtained unprecedented evaluation metrics and could prove to be extremely useful when it comes to fast and accurate detection of the disease.