摘要
着重介绍受试者工作特征(Receiver Operating Characteristic,ROC)曲线的理论,以及在诊断试验准确性评价中的应用。通过对五种乙型肝炎病毒表面抗原诊断试剂评价的实例,简要说明ROC曲线下面积(Area Under Curve,AUC)、部分ROCAUC和固定特异度时灵敏度的计算方法及其意义,以及这些指标相互比较的Bootstrap方法。
The theory and application of ROC curve are briefly introduced in this paper. Three main ROC curve related indices, area under curve （AUC）, partial area under curve （pAUC）, and sensitivity at a particular false-positive-rate （1-specificity） [Se（FPR=e）], are used to analyze a real data of five hepatitis B surface agent kits with calculated steps and explanation, including the bootstrap method for statistical inference of these indices.
出处
《中国疫苗和免疫》
CAS
2009年第5期462-466,共5页
Chinese Journal of Vaccines and Immunization
