COVID-19 Classification using CT Scan Images with Resize-MobileNet

ABSTRACT

Coronavirus disease 2019 broke out in early 2020 and quickly spread to over 200 countries, leading to a severe health crisis for people all over the world. In high-risk areas of the epidemic, the shortage of testing reagents and medical facilities have become essential factors restricting the treatment of COVID-19 patients. Computed tomography (CT) has helped doctors make medical diagnoses in many areas as a vital technology in medical field. At present, due to personal privacy issues, it isn't easy to compare different networks because they are all conducted on different data sets, using other metrics, and can not make good use of high-resolution CT images. Based on iCTCF's public data set, 4000 photos from 61 patients are used to propose a network of high-resolution inputs for diagnosing disease using lung CT images of COVID-19 patients. Our work makes better results than traditional image classification methods in limited data sets, contributing to the advancement of deep neural networks in the field of COVID-19CT image recognition. © 2021 IEEE.