ABSTRACT
Detection of COVID-19 has been a very active field of research with thousands of papers published after outbreaks of COVID-19 in the world. Computer-Aided Design (CAD) based studies have a significant role in the medical field thanks to rapidly developing technology. To help radiologists speed up the diagnostic process, CAD with convolutional neural networks (CNN) can be used as decision support mechanisms. Furthermore, CNN has the power to learn various image features automatically, and it may offer an effective way for COVID-19 detection. In this paper, we propose a CNN design for COVID-19 detection. We used a data set of X-ray images collected from two publicly available sources. This data set consists of 400 images of which 200 are COVID-19 and 200 are healthy. First, we preprocessed all data sets and then divided them by randomly allocating 70 % for training and 30 % for the test. We obtained the accuracy, specificity, and sensitivity rate of our model as 96.11%, 98.89 %, and 93.33 %, respectively. © 2022 IEEE. All rights reserved