Multiclass Classification of Covid-19 CT Scan Images With VGG-16 Architecture Using Transfer Learning System

Abstract

COVID-19 is a respiratory disease caused by the coronavirus. The most common test technique used today for COVID-19 diagnosis is real-time reverse transcription-polymerase chain reaction (RT-PCR). However, compared to RT-PCR, radiological imaging such as X-rays and computer tomography (CT) may be a more precise, useful, and faster technology for COVID-19 classification. X-rays are more accessible because they are widely available in all hospitals in the world and are cheaper than CT scans, but the classification of COVID-19 using CT scan images is more sensitive than X-rays. Therefore, CT scan images can be used for the early detection of COVID-19 patients. One of them is using the deep learning method. In this study, a CNN algorithm with a VGG-16 architecture will be selected to classify COVID-19, intermediate, and non-COVID CT scan images using 2481 image datasets. First, pre-processing is done by resizing the image, converting the image channel into RGB, and dividing the dataset into a training dataset and a testing dataset. Then, the convolution process is continued by utilizing the pre-trained VGG-16 model from ImageNet. The results of testing the data with 97% accuracy were obtained. It is concluded that the model used to classify COVID-19, intermediate, and non-COVID CT scan images is effective and produces good results.