Classification of COVID19 X-ray images based on transfer learning INCEPTIONV3 deep learning model

ABSTRACT

The World Health Organization (WHO) has recently announced the novel coronavirus 2019 as a pandemic. Many preventative plans and non-pharmaceutical efforts have emerged and been in use to manage and control the spread of the disease which includes infection control, proper isolation of patients, and social distancing. The main test used to confirm a COVID-19 case is the RT-PCR test. However, this approach needs analysis time and specimen collection. Therefore, the importance of medical imaging is increased to screen COVID-19 cases. Hence radiology has a pivotal role in managing COVID-19 infection using CT scans and chest X-ray (CXR) throughout the screening, diagnosis, and prognostication processes of the disease. In this paper, a new model using the transfer learning method and InceptionV3 algorithm has been presented to classify the X-ray images into COVID-19, Normal, and Pneumonia classes. The experimental results show that the proposed model achieved 98% Accuracy on the test set for classifying the images from the 3 different classes. © 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.