Evaluating Median Accuracy of ResNet50 and VGG16 Models in COVID-19 Detection

Abstract

The increasing number of Covid-19 cases and the lack of reliable, quick-to-use testing tools herald a new era in X-ray analysis employing deep learning methods. The Covid-19 virus's emergence poses a threat to human existence. Therefore, it will take time to develop a quick and accurate method of identifying the Covid-19 virus in patients. The reference method is the conventional RT-PCR technique. The goal of this research is to create an automated system that uses CNN models such as Resnet50, VGG16, and Grad-CAM to analyze X-ray images in order to provide a reliable and effective method of diagnosing Covid-19 infection. The created models use image processing techniques to pre-process the X-ray picture. Afterwards, deep learning is used to classify the images after they have been segmented and transformed. The CNN model that is being used provides strong classification accuracy and shows the location in the lung where the disease is attacked, even for a normal person, we can anticipate the likelihood of where the COVID may affect them. Our model utilizes a convolution neural network that is trained on the standard COVID-19 Radiography Dataset.