AUCs (95%CIs) from ROC curves for classifying tumor presence status based on tumor presence confidence scores given by group and reader for all tumors (n = 107).

* Comparisons between interpreters: ACPPD-Gd, p = 0.04; gadobutrol, p = 0.003; non-contrast, p = 0.0003. AUCs (95%CIs) from ROC curves for classifying tumor presence status based on tumor presence confidence scores given by group and reader for all tumors (n = 107).阅片者间的对比：ACPPD-Gd，p=0.04；钆布醇（gadobutrol），p=0.003；无对比增强组，p=0.0003。针对全部107例肿瘤，基于各分组与阅片者给出的肿瘤存在置信度评分，通过受试者工作特征（Receiver Operating Characteristic, ROC）曲线计算得到的肿瘤存在与否分类的曲线下面积（Area Under Curve, AUC）与95%置信区间（95% Confidence Intervals, 95%CIs）。