NCOVID-19 Detection with Frontal Chest X-Ray Scans Using Deep Learning Technique

Abstract

The novel coronavirus (nCOVID-19) epidemic, which has infected millions of people and killed a large number of people throughout the world, was immediately declared a global emergency. This disease is caused by the highly infectious virus ‘Severe Acute Respiratory Syndrome CoronaVirus-2 (SARS-COV-2)’, in which infection with cryptogenic organizing pneumonia can be spread even from asymptotic patients during the incubation stage. According to the expert's perspective, the virus mainly affects the human respiratory tract, causing severe bronchopneumonia with fever, dyspnea, dry cough, tiredness, and respiratory failure. For identifying nCOVID-19, the usual ‘Reverse Transcription-Polymerase Chain Reaction (RT-PCR)’ clinical confirmatory test is manual, complex, and time-consuming. One of the key reasons for the need for a faster and easier method of diagnosing infected individuals to be separated and cared for is the limited availability of domain experts and test kits in hospitals. An automatic screening method may act as a second opinion for medical practitioners to rapidly identify infected individuals who require immediate isolation. Additional clinical confirmation is necessary when there is a transitory spike in the number of infected patients. We are proposing a deep learning technique to detect the nCovid-19 using frontal Chest X-ray scans.