Abstract
AbstractCOVID-19 had a huge impact on patients and medical systems all around the world. Computed tomography (CT) images can effectively complement the reverse transcription-polymerase chain reaction testing (RT-PCR) and offer results much faster than RT-PCR test which assists to prevent spread of COVID-19. Various deep learning models have been recently proposed for COVID-19 screening in CT scans as a tool to automate and help the diagnosis, but consisting of some benefits and limitations. Some of the reasons for this are: (i) training the data with largely unbalanced dataset and (ii) training the models with datasets having all similar CT images which leads to overfitting. In this work, we proposed a method to use multiple models to classify COVID-19 positive or negative which are trained using transfer learning techniques. In addition to classifying, if a person is COVID-19 positive or negative, we have also calculated the high-resolution computed tomography (HRCT) score or CT score to find the severity of infection with the help of image segmentation techniques, which assist in identifying the preliminary prognosis of the patient, and take necessary preventive measures.
Publisher
Springer Science and Business Media LLC
Subject
General Earth and Planetary Sciences,General Engineering,General Environmental Science
Reference40 articles.
1. He K, Zhang X, Ren S, Sun J (2015) Deep residual learning for image recognition. arXiv 2015. arXiv preprint arXiv:1512.03385
2. Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556
3. Bansal M, Kumar M, Sachdeva M, Mittal A (2021) Transfer learning for image classification using VGG19: Caltech-101 image data set. J Ambient Intell Humaniz Comput 1–12
4. Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2015). Rethinking the inception architecture for computer vision. CoRR http://arxiv.org/abs/1512.00567
5. Chollet F (2016) Xception: Deep learning with depthwise separable convolutions, CoRR abs/1610.02357. URL http://arxiv.org/abs/1610.02357