ABSTRACT
The world is affected by an existential global health crisis called the COVID-19 pandemic. Countries like the United States, India and Russia are still having and gaining positive COVID cases, which results in the deaths of hundreds and thousands of people. Admitting the actuality that there are several vaccinations on the market at the moment, positive cases continue to rise. Consequently, there is a critical need for quick infection detection with clear visualization so that a suspected COVID-19 patient can be spared. Tests, namely Polymerase chain reaction (PCR), Lateral flow tests (LFTs), need to send to a laboratory for examination, so patients may have to wait for a few days to get their results, but still, the final results aren't accurate. CT Scan pictures are a commonly utilized imaging modality among previously existing, low-cost and widely available resources, but deep learning approaches have attained state-of-the-art performance in computer-aided medical diagnosis. The goal of our paper is to employ CNN (Convolutional Neural Network) and find whether the patient has COVID-19 or not by using CT scan pictures. The suggested method employs convolutional neural networks as part of its deep learning techniques. COVID19 was diagnosed using the CNN model with various filters, and they achieved accuracy with 85.34%, 87.46% and 88.15%, respectively. Doctors can employ COVID-19 automated diagnosis using CT scan pictures as a quick and effective technique to detect COVID-19 © 2022 IEEE.