Value of multiparameter quantitative CT in evaluating the prognosis of COVID-19

Objective: To investigate the value of multi-parameter quantitative CT combining with HRCT in evaluating the prognosis of coronavirus disease 2019 (COVID-19) Methods: Sixty-nine cases were included in this retrospective study All of the cases were primarily diagnosed with common type of COVID-19 when admitted to hospital According to the clinical prognosis and the changes of image appearance, three groups were divided Group A with decreasing extent of lesions (21 cases) and group B with increasing extent of lesions (25 cases) were both common type of COVID-19 without converting into severe type Group C (23 cases) converted common type into severe type during hospitalization Multiple CT parameters among the groups were compared Results: The changes of the multiple parameters of CT after treatment, such as the total volume of lesions, GGO volume, and the volume of the consolidation and MLA, had a significant difference among the three groups When the cut-off values for the increasing total volume, GGO volume, consolidation volume, and MLA were determined to be 231 46 cm3, 168 58 cm3, 74 46 cm3, and 57 Hu, the sensitivity (95 7%, 82 6%, 73 9%, and 78 3%) and specificity (91 3%, 91 3%, 93 5%, and 87%) of the diagnosis of common type converting into severe type were optimal and the areas under ROC curve were 0 961, 0 914, 0 885, and 0 885 Conclusion: The changes of the volumes of total lesions, GGO and consolidation and the change of MLA between follow-up CT and first CT are important prognostic factors in patients with COVID-19 Common type COVID-19 patients with total volume of lesions increasing >231 46 cm3, GGO volume increasing >168 58 cm3, consolidation volume increasing >74 46 cm3, or MLA increasing >57 Hu are more likely to convert into severe type The multiparameter quantitative CT combined with HRCT has an effect on evaluating the prognosis of COVID-19 © 2021, Editorial Board of Medical Journal of Wuhan University All right reserved