Deep learning based prediction of COVID-19 virus using chest X-Ray

Abstract

PCR (Polymerase Chain Reaction) tests are used as a standard test for the COVID-19, which is having a lot of issues. For finding the exact results using the PCR testing technique, multiple testing is needed, which is a time-consuming process. As it is profoundly dangerous when contaminated ones are not isolated from the others before time, they can affect others as well. In today’s time, there is an urgency for the development of an automatic detection system as a substitute for the expeditious treatment of COVID-19 affected patients. This research aim is to increase the testing efficiency of COVID-19 by predicting the presence of COVID-19 by utilizing the Chest X-Ray images. In this research, CNN (abbreviated as Convolutional Neural Network) algorithm of Deep Learning is used to predict and analyze the impact of COVID-19 on the human body by building a deep neural network of 11 convolutional layers. The dataset for this research is taken from Kaggle repository. The dataset contains the Chest X-Ray images in Posteroanterior (PA) view) of normal as well as Covid affected persons. The model is created, and the results have been evaluated by using the various evaluation metrics, i.e., Recall, F1-Score, and Precision. The results obtained from the validation testing, the Recall, F1-Score, and Precision, are 0.94, 0.97, and 1.0, respectively, which indicates the very high accuracy of the proposed model.