Automatic X-ray COVID-19 Lung Image Classification System based on Multi-Level Thresholding and Support Vector Machine

Aboul Ella Hassanien; Lamia Nabil Mahdy; Kadry Ali Ezzat; Haytham H. Elmousalami; Hassan Aboul Ella

A radiological imaging of confirmed COVID-19 infected patient typically shows either patchy or diffuse asymmetric airspace opacities, similar to the SARS and MERS associating pneumonia. A bilateral lung affection at first few days of infection in 98% of cases was reported at the first report of patients with COVID-19, with two different patterns according to intensive care unit (ICU) association or not and it is described as follow, patients inside intensive care units (ICUs) showed consolidation pattern while patients outside (ICUs) showed ground-glass predominant pattern. Also a 21 confirmed COVID-19 cases investigatory study showed chest radiological images with lung affections in 86% of patients, with a majority of 89% having bilateral lung affections [28] . Peripheral lung affection with consolidation or multifocal ground-glass predominant opacities was reported in 29% and 57% of cases, respectively. As well, a COVID-19 confirmed seven people family cluster chest radiological images showed bilateral patchy ground-glass predominant opacities with greater involvement of older family member's lung than those young members [27, 29] . Although the imaging features closely similar and overlapping those associating of SARS and MERS, the bilateral lungs involvement on initial imaging is more likely to be seen with COVID-19; as those associating SARS and MERS are more predominantly unilateral.