Deep Learning Approach for Diagnosing COVID-19 in CT Images

Abstract

AbstractThe Novel Coronavirus disease 2019 (COVID-19) is the briskly transmittable virus caused by severe acute respiratory syndrome Coronavirus 2 (SARS-CoV-2). The initial case was identified in Wuhan city, China in December 2019. The disease expanded worldwide, leading to an ongoing pandemic. As reported by the World Health Organization, COVID-19 has already spread to more than 160 million people worldwide and is responsible for 35,24,060 deaths. Computer-Aided Diagnosis (CAD) is a non-invasive procedure that performs the detection of COVID-19 using deep learning will assist to recognize this virus in the initial stages so that it will register in elevating the chances of early recovery of the COVID-19 victims worldwide. Here CT scan images were used and the insufficient standard datasets for COVID-19 particularly in lung CT Images are the prior reason for this study. In this proposed approach, a deep convolutional neural network (CNN) based model InceptionV3 is preferred for the exploration to detect the Coronavirus-affected patients using lung CT scan images. The end results show that InceptionV3 is the most suitable deep learning model to notice or to identify the COVID-19 from the CT scan dataset with a testing accuracy of 99.19%, precision of 100%, recall of 95%, and f1 score of 97.22%.KeywordsComputer-Aided DiagnosisInceptionV3COVID-19Deep CNNDeep LearningCT photographsMedical Image Analysis