An Unsupervised Approach for COVID-19 Detection using Chest CT Images

Abstract

In this work we assume we have data of normal (healthy) and general pneumonia infected subjects. Based on this knowledge, can we predict COVID-19 infection as abnormal health condition? If so what is the severity of infection? This study will help identify new variants, mutations of COVID-19 as abnormal health conditions. In long term, our study can help identify new infections altogether based on data of healthy subjects itself.