初探新冠肺炎早期影像特征评分在诊断疑似病例中的价值

A preliminary study on the value of early image feature score in the diagnosis of suspected cases of COVID-19

目的研究新冠肺炎早期影像特征评分在诊断疑似病例中的价值。方法回顾性分析120例新冠肺炎疑似病例影像特征,由两组影像医师(a组:高年资组;b组:低年资组),依次用经验法及评分法对120例新冠肺炎疑似病例影像特征进行分析诊断,运用Kappa检验和ROC曲线进行各组数据比较和分析。结果(1)经验法,两组影像医师对120例疑似病例诊断一致性相差较远。(2)评分法,两组影像医师对120例疑似病例诊断高度一致。通过ROC曲线分析,当放射医师评分>5.5时,应高度警惕新冠肺炎疑似病例为新型冠状病毒肺炎(COVID-19)。结论新冠肺炎早期影像特征评分法对诊断新冠肺炎疑似病例时有一定的价值,其结合流行病学史及临床表现,可进行全面辩证分析,从而增大诊断的客观准确性。

Objective To investigate the value of early image feature score in the diagnosis of suspected cases of Corona Virus Disease 2019(COVID-19).Methods Image features of 120 suspected COVID-19 cases were retrospectively analyzed and diagnosed by two groups of imaging physicians(group a:senior physicians,group b:junior physicians)by means of specific empirical method and scoring method.They used Kappa test and ROC curve to compare and analyze the data of each group.Results(1)Empirical method showed that the diagnoses consistency of two groups of imaging physicians in 120 suspected cases were far from the same.(2)Scoring method showed that the diagnoses of two groups were highly consistent in the 120 suspected cases.According to ROC curve analysis,when the radiologist score was higher than 5.5,physicians should be on high alert for the suspected cases of COVID-19.Conclusion Early image feature score is of certain value in the diagnosis of suspected COVID-19 cases.Combined with epidemiological history and clinical manifestations,it can be used for comprehensive dialectical analysis,so as to increase the objective accuracy of diagnosis.