COVIHunt: An Intelligent CNN-Based COVID-19 Detection Using CXR Imaging

ABSTRACT

Recently, the individuals are under lockdown and limited mobility due to the random spreading of the COVID-19, i.e., coronavirus disease - 2019, worldwide as well as pandemic declared by the World Health Organization (WHO). RT-PCR, i.e., reverse transcriptase-polymerase chain reaction, tests that can detect the RNA from nasopharyngeal swabs have become the norm to allow people to travel within the nation and also to international destinations. This test is people-intensive, i.e., it involves a person collecting the sample, needs transportation with strict precautionary measures, and a lab technician to perform the test which may take up to 2 days to get the results. There is a lot of inconvenience to the people due to this process. Alternatively, X-Ray images have been used primarily by physicians to detect COVID-19 and its severity. Detection of COVID-19 through X-Ray can act as a safe, faster, and alternative method to RT-PCR tests. This method uses a Convolutional Neural Network (CNN) to classify the X-Ray scans into two categories, i.e., COVID-19 positive and negative. In this paper, a novel method named COVIHunt an intelligent CNN-based COVID-19 detection technique using CXR imaging, is proposed for binary classification. From experiments, it is observed that the proposed work outperforms in comparison with other existing techniques. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.