Classification of Coronavirus Images using Shrunken Features

Saban Ozturk; Umut Ozkaya; Mucahid Barstugan

The Coronavirus (COVID- 19) , which appeared towards the end of 2019, caused a global epidemic problem that could spread quickly from the individual to the individual in the community. According to the World Health Organization (WHO) data, the rate of catching COVID-19 in China is between 16-21%, and the mortality rate is 2-3% [1] . For this reason, people need to take the necessary quarantine measures and implement protection procedures to deal with COVID-19. The clinical symptom must be present, and also positive X-ray and CT images must be included with the positive pathological test to be able to diagnose COVID-19. Fever, cough, and shortness of breath are frequently seen as clinical symptoms [2] . Besides, positive findings should be obtained in X-ray and CT images with these symptoms. Another diagnostic method of COVID-19 is to examine the RNA sequence of the virus. However, this method is not a very efficient technique due to a long time of diagnosis. Also, its accurate diagnosis rate varies between 30-50%. Thus, diagnosis tests should be repeated in lots of time [3] . Radiological imaging techniques are essential for the detection of COVID- 19 . In X-rays and Computer Tomography (CT) images, the COVID-19 virus shows the same features in the early and late stages. Although it shows a circular distribution within an image, it may exhibit similar characteristics with other viral epidemic lung diseases [4] . This makes it challenging to detect COVID- 19 from other viral cases of pneumonia.