Deep Convolutional Neural Network Architecture to Detect COVID-19 from Chest X-Ray Images

Abstract

Today, the world is living in a time of epidemic diseases that spread unnaturally and infect and kill millions of people worldwide. The COVID-19 virus, which is one of the most well-known epidemic diseases currently spreading, has killed more than six million people as of May 2022. The World Health Organization (WHO) declared the 2019 coronavirus disease (COVID-19) after an outbreak of SARS-CoV-2 infection. COVID-19 is a severe and potentially fatal respiratory disease caused by the SARS-CoV-2 virus, which was first noticed at the end of 2019 in Wuhan city. Artificial intelligence plays a meaningful role in analyzing medical images and giving accurate results that serve healthcare workers, especially X-ray images, which are complex images in their interpretation. In this article, two deep convolutional neural network (DCNN) classifiers, such as Inception-v2 and VGG-16, are utilized to detect COVID-19 from a set of chest X-ray images. The dataset for this article was collected from the Kaggle platform (COVID-19 Radiography Database) and consists of images of positive and healthy people. This article concludes that the most suitable performance is the Inception-v2 classifier, which has achieved an accuracy of 97% in comparison to the VGG-16 classifier, which has achieved an accuracy of 93%.