Coronavirus Disease 2019 (COVID-19) Candidate Chest CT Features: A Systematic Review of Extracted Imaging Features from 7571 Individuals

Since the outbreak of Coronavirus Disease 2019 (COVID-19) causing novel coronavirus (2019-nCoV)-infected pneumonia (NCIP), over 45 million affected cases have been reported worldwide. Many patients with COVID-19 have involvement of their respiratory system. According to studies in the radiology literature, chest computed tomography (CT) is recommended in suspected cases for initial detection, evaluating the disease progression and monitoring the response to therapy. The aim of this article is to review the most frequently reported imaging features in COVID-19 patients in order to provide a reliable insight into expected CT imaging manifestations in patients with positive reverse-transcription polymerase chain reaction (RT-PCR) test results, and also for the initial detection of patients with suspicious clinical presentation whose RT-PCR test results are false negative. A total of 60 out of 173 initial COVID-19 studies, comprising 7571 individuals, were identified by searching PubMed database for articles published between the months of January and June 2020. The data of these studies were related to patients from China, Japan, Italy, USA, Iran and Singapore. Among 40 reported features, presence of ground glass opacities (GGO), consolidation, bilateral lung involvement and peripheral distribution are the most frequently observed ones, reported in 100%, 91.7%, 85%, and 83.3% of articles, respectively. In a similar way, we extracted CT imaging studies of similar pulmonary syndromes outbreaks caused by other strains of coronavirus family: Middle East Respiratory Syndrome (MERS) and Severe Acute Respiratory Syndrome (SARS). For MERS and SARS, 2 out of 21 and 5 out of 153 initially retrieved studies had CT findings, respectively. Herein, we have indicated the most common coronavirus family related and COVID-19 specific features. Presence of GGO, consolidation, bilateral lung involvement and peripheral distribution were the features reported in at least 83% of COVID-19 articles, while air bronchogram, multi-lobe involvement and linear opacity were the three potential COVID-19 specific CT imaging findings. This is necessary to recognize the most promising imaging features for diagnosis and follow-up of patients with COVID-19. Furthermore, we identified co-existed CT imaging features.