Deep Learning Significance in COVID-19 Detection: A Review of CNN Architectures and Medical Image Analysis

Abstract

Beginning in Wuhan, China, the corona-virus disease 2019 (COVID-19) epidemic spread in early December and has subsequently extended elsewhere. Radiologists are able to determine whether radiographic abnormalities are present in the COVID-19 patients by looking at their chest X-ray (CXR) images. This paper examines the application of Convolutional Neural Networks (CNN) in the detection of COVID-19, emphasizing the rationale behind their implementation in medical image processing. In addition, it investigates specific CNN models that includes ResNet50, DenseNet121, and VGG16, for the purpose of detecting COVID-19. Furthermore, the investigation encompasses image enhancement methodologies and a variety of COVID-19 datasets, including the COVIDx, the COVID-19 Radiography, and the SARS-CoV-2 CT Scan dataset.