摘要
目的比较ADNEX模型、简单规则风险估计模型及恶性风险指数(RMI)对卵巢良恶性肿瘤的诊断价值。方法回顾性分析286例卵巢肿瘤患者的术前超声检查图像,分别以ADNEX模型、简单规则风险估计模型及RMI评估卵巢肿瘤的良恶性。以病理结果为金标准,计算并比较3种方法的敏感度、特异度,绘制ROC曲线,获得曲线下面积。结果286例卵巢肿瘤患者中,良性肿瘤142例,恶性肿瘤144例。ADNEX模型、简单规则风险估计模型、RMI诊断卵巢良恶性肿瘤的敏感度分别为83.33%(120/144)、80.56%(116/144)、65.97%(95/144),特异度分别为89.44%(127/142)、92.96%(132/142)、90.14%(128/142)。ADNEX模型与简单规则风险估计模型间敏感度及特异度差异均无统计学意义(χ~2=0.352、1.784,P=0.554、0.182),ADNEX模型及简单规则风险估计模型的敏感度均高于RMI(χ~2=16.691、7.533,P均<0.001),而特异度与RMI间差异均无统计学意义(χ~2=0、0.561,P=1.000、0.454)。ADNEX模型、简单规则风险估计模型及RMI鉴别卵巢肿瘤良恶性的AUC分别为0.864、0.868和0.788(P均<0.001)。结论 ADNEX模型及简单规则风险估计模型鉴别卵巢肿瘤良恶性的效能在一定程度上优于RMI。
Objective To compare the value of ADNEX model,simple rules risk model and the risk of malignancy index(RMI)in diagnosis of benign and malignant ovarian tumors.Methods The preoperative ultrasonic images of 286 patients with ovarian tumors were retrospectively analyzed.ADNEX model,simple rules risk model and RMI were used to differentiate benign and malignant ovarian tumors.Taken histopathological results after surgery as golden standards,the sensitivity and specificity were calculated and compared among 3 methods.ROC curve was used to obtain the area under the curves.Results Among 286 ovarian tumors,142 were benign and 144 were malignant.The sensitivity of ADNEX model,simple rules risk model and RMI was 83.33%(120/144),80.56%(116/144)and 65.97%(95/144),respectively,while the specificity was 89.44%(127/142),92.96%(132/142)and 90.14%(128/142),respectively.There was no statistical difference of sensitivity nor specificity between ADNEX model and simple rules risk model(χ^2=0.352,1.784,P=0.554,0.182).The sensitivity of ADNEX model and simple rules risk model was higher than that of RMI(χ^2=16.691,7.533,respectively,both P<0.001),while there was no statistical difference of specificity(χ^2=0,0.561,P=1,0.454).The AUC of ADNEX model,simple rules risk model and RMI was 0.864,0.868 and 0.788,respectively(all P<0.001).Conclusion ADNEX model and simple rules risk model are better than RMI in differentiating benign and malignant ovarian tumors.
作者
和平
吴青青
孙丽娟
王晶晶
王莉
韩吉晶
张铁娟
HE Ping;WU Qingqing;SUN Lijuan;WANG Jingjing;WANG Li;HAN Jijing;ZHANG Tiejuan(Department of Ultrasound,Beijing Obstetrics and Gynecology Hospital,Capital Medical University,Beijing 100026,China)
出处
《中国医学影像技术》
CSCD
北大核心
2019年第1期104-107,共4页
Chinese Journal of Medical Imaging Technology
基金
国家重点研发计划(2016YFC1000104)
北京市医院管理局"登峰"计划专项经费(DFL20151302)