Accurate Prediction of COVID-19 using Chest X-Ray Images through Deep Feature Learning model with SMOTE and Machine Learning Classifiers

Rahul Kumar; Ridhi Arora; Vipul Bansal; Vinodh J Sahayasheela; Himanshu Buckchash; Javed Imran; Narayanan Narayanan; Ganesh N Pandian; Balasubramanian Raman

Given the deadly and debilitating condition that it is, the entire world now is at high alert with thousands of cases being identified each day. Going by WHO statistics, as of now, there have been about 1.84 million identified cases with over 1, 13, 329 deaths 6 . Country-wise (top 10 India) statistical COVID-19 cases as of April 12, 2020, presented in Table 1 7 . According to WHO 8 , geographical statistics of the confirmed COVID-19 cases have registered as of April 12, 2020 as depicted in the Figure 1 . International travel history and close contact with the infected have been identified as the causes of the worldwide spread of the disease. An enormous amount of effort is being put into developing vaccines and drugs to help treat the infection 9, 10 . Speaking of India, the country currently stands at 9, 205 confirmed cases and 331 deaths. There has been a 21-day complete lockdown in India as a measure to prevent further transmission, but whether India is coping up with the disease or not continues to be a big question. While these small numbers of confirmed and deceased cases may bring in a sigh of relief, they also challenge the state of medical testing and diagnosis in India 11 . Whether these numbers are low or they are low because of insufficient testing remains unanswered 12 .