ABSTRACT
In the field of medical imaging, deep learning techniques have already proven to be quite a success.The global population's health and well being continue to be severely impacted by the Coronavirus Disease 2019 (COVID-19) pandemic, healthcare systems are unable to examine and diagnose patients as soon as they ought to. The various post COVID complications which may have been dealt in a better way if the virus was detected at an earlier stage and given appropriate clinical support. Chest radiography imaging is essential for detecting and tracking COVID-19 because of its effects on pulmonary tissues. Chest X-Ray(CXR) imaging is even more readily available than chest computed tomography(CT) imaging, especially in developing countries where CT scanners are too costly due to high equipment and maintenance costs. In this work we propose a very lightweight convolutional neural network (CNN), in which the chest X-Ray samples comprising of COVID-19, Non-COVID and Normal cases are analyzed without any human intervention. Our model gives comparable accuracy to other COVID-19 detection models proposed earlier while having significantly fewer parameters than them, which makes our model optimal for deployment on machines with low computing power. © 2022 IEEE.