COVID-19 Detection in CT Images using Deep Transfer Learning

Abstract

Confronting the COVID-19 pandemic introduced by newest corona virus, SARS-CoV-2, is one of the human species' most influential problems today. The fast identification and isolation of infected patients is a crucial factor in slowing down the spread of the virus. The Reverse Transcription Polymerase Chain Reaction (RT-PCR) process, one of the basic methods for COVID-19 recognition, is time-consuming in addition short-lived due to the pandemic. Deep learning applied to patients' CT images has given away hopeful results popular the identification of COVID-19 in this context. The powerful net family of CNN models using CT images to perform COVID-19 recognition is suggested in this article by VGG-16. As a consequence, COVID-19 detection was proposed as a VGG-16 model with an overall accuracy of 98.33 percent. We assume that, both in terms of productivity and efficiency, the published figures reflect modern outcomes.

Keywords

• VGG-16
• Deep learning
• CT images
• Convolution neural networks (CNN)