To investigate the correlations of initial lab and imaging findings in COVID-19 patients of different clinical types.
We retrospective analyzed patients confirmed with COVID-19 in the Fifth Medical Center of the People’s Liberation Army (PLA) General Hospital between February to April 2020, selected a total of 58 (N) patients with lab and imaging examinations that met the study criteria, using Artificial intelligence (AI) software to calculate the percentage of COVID-19 lesions in the volume of the whole lung, then the correlations of general information, initial chest CT examination after admission and laboratory examinations were analyzed.
The 58 (N) COVID-19 patients were divided into mild group [41(n) cases]: and severe group [17(n) cases]: according to patient’s condition. CT findings of the severe group and mild group mainly included single or multiple ground glass opacity (GGO), with lesions mainly distributed in the periphery of lungs or GGO mixed with consolidation, with lesions involved in peripheral and central areas of both lungs, accompanied other signs. A significant difference in CRP, IL-6, D-D, GGT was observed between the two groups (p < 0.05). The ratios regarding lymphocyte abnormality and neutrophil abnormality in the severe group were higher than those in the mild group (p < 0.05).
The CT features at initial diagnosis of COVID-19 were mainly characterized by multiple GGO with or without partial consolidation in both lungs, with the lesions mainly distributed at the subpleural regions. Some lab test indexes were correlated with the clinical types of COVID-19.
Artificial intelligence; Clinical characteristics; Computed tomography; Coronavirus Disease 2019; Imaging features.