Classification of Coronavirus Images using Shrunken Features

Saban Ozturk; Umut Ozkaya; Mucahid Barstugan

Machine learning techniques, which are a sub-branch of the field of artificial intelligence, are frequently used for medical applications in the concept of feature extraction and image analysis. Machine learning methods are used in the diagnosis of viral pneumonia, especially in X-ray and CT images, tumor diagnosis, and cystoscopic image analysis [5] . Viral pathogenic patterns contain several features in X-ray and CT images [6] . COVID-19 shows an irregular distribution and shading within the image [7] . When the X-ray and CT studies in the literature are examined, it is seen that hand-crafted, and recently CNN-based methods are used [8, 9] . Studies with hand-crafted features are generally seen in studies performed before the CNN method. Studies with hand-crafted features are generally seen in studies performed before the CNN method. These studies are highly dependent on the results produced by the methods chosen based on the researcher's experience. In addition, it is not always possible to achieve the same performance for different datasets. These methods cover many techniques from edge detection algorithms to GLRLM and SFTA methods.