Potential Diagnostic Application Of A Novel Deep Learning- Based Approach For Covid-19

This paper presents the COVID-MAH-CT dataset, which comprises chest CT scans of 133 patients who tested positive for COVID-19 and were referred to training-therapeutic hospitals in northern Iran. The dataset was selected purposefully during four different waves of the COVID-19 epidemic in Iran. The CT scan images used for the diagnosis and treatment of patients were captured by CT scan imaging devices of the same model and visualized using the same software version. Three radiologists examined the images and classified them as positive or negative for COVID-19. Positive radiological findings are in the form of Ground glass opacities (GGOs). The paper also presents the results of applying MASERes to a COVID-19 dataset other than COVID-MAH-CT, and presents the results of an ablation study on MASERes.