A novel machine learning‐based analytical framework for automatic detection of<scp>COVID</scp>‐19 using chest<scp>X‐ray</scp>images

Considering the prevailing scenario of COVID-19 pandemic, early detection disease is an important and crucial step in management. Early correct treatment may limit progression to severe levels prevent deaths. In addition, isolation infected patients will lead control transmission rate possibly reduce stress on present healthcare system. Currently, most common reliable testing method available for diagnosis real-time reverse transcription-polymerase chain reaction (rRT-PCR) test. However, chest radiological (X-ray) imaging can be used as alternate rRT-PCR test, symptoms investigated by critical examination patient's scans. work, a novel machine learning (ML)-based analytical framework developed automatic using X-ray (CXR) images plausible patients. The designed, trained, validated identify four classes CXR namely, healthy, bacterial pneumonia, viral COVID-19. experimental results pose proposed potential candidate images, with training, validation, accuracy 92.4%, 88.24%, 87.13%, respectively, four-class classification. comparative analysis demonstrates better capabilities along other types pneumonia.