CT Diagnosis of COVID-19 Based on the Imaging Signs of Secondary
Pulmonary Lobular
Abstract
Objectives To study the CT diagnosis of COVID-19 that was based on the
imaging features of secondary pulmonary lobular structure. Materials and
methods After institutional review board approval and informed consent,
139 cases with suspected COVID-19 who undergone CT were retrospectively
analyzed. Based on the anatomical structure of pulmonary lobule, the
imaging features of secondary pulmonary lobular structure on CT images
was defined and built simulation diagrams serve as a unified standard.
The imaging performance of confirmed 103 cases was evaluated for
diagnosis of COVID-19. Results Of these 139 cases, 103 cases were
confirmed for COVID-19, 36 cases were confirmed for other diseases.The
sensitivity, specificity and accuracy of CT are 83.33%, 20.39%,
80.58%,Of the 103 COVID-19 cases, 82(79.61%) cases were typical cases
and 21(20.39%)cases were atypical cases. 65(63.11% ) cases were in
the early or advanced stage, 26(25.24%) cases were in the severe or
recovery stage, and 12(11.65%)were in multiple stages. The scores of
pulmonary lobular structure in typical COVID-19 (3.42±1.25) were
significant different compared with atypical COVID-19 (1.51±1.34,
P<0.001). The inter-observer agreement for the score of
secondary pulmonary lobular structure was excellent (kappa = 0.863,
P<0.05). Conclusion To recognize the CT common imaging signs
of secondary pulmonary lobular structure in COVID-19 can deepen the
understanding of COVID-19 from the basis of image anatomy, and screen
suspected cases and provide imaging basis for diagnosis and differential
diagnosis.