| 摘要: |
| 目的 对比计算机辅助诊断系统(CAD)基于不同指南诊断甲状腺结节良恶性的效能。 方法 多中心、前瞻性分析2019年1月至2019年9月就诊于四所联盟医院的453例患者(共521例甲状腺结节),所有结节均经手术或超声引导下细针穿刺病理证实。以病理为金标准,建立受试者工作特征曲线(ROC曲线),并得到曲线下面积(AUC),cut-off值及其对应的的敏感度、特异度、阳性预测值(PPV)、阴性预测值(NPV)。组间AUC对比采用Z检验,组间敏感度、特异度对比采用McNemar检验。 结果 KSThR共识的AUC、敏感度高于其他7个指南的AUC(P均＜0.01),BTA指南的特异度高于其他7个指南(P均＜0.01)。 结论 KSThR共识具有最高的AUC、敏感度,BTA指南具有最高的特异度。 |
| 关键词: 甲状腺结节 超声 计算机辅助诊断 |
| DOI： |
| 投稿时间：2019-10-25修订日期：2019-10-25 |
| 基金项目:泉州市科技计划项目(2019C076R)、教育部泉州医学高等专科学校母婴健康服务应用技术协同创新中心经费资助、福建省临床重点专科建设项目经费资助 |
|
| Comparison of diagnostic efficacy of computer-aided diagnosis system based on different guidelines for the diagnosis of benign and malignant thyroid nodules |
| YE Fengying,LYU Guorong,SU Qichen,LI Shangqing,YOU Jianhong,WANG Kangjian,CAI Mingli |
| (Department of ultrasound,the Second Affiliated Hospital of Fujian Medical University) |
| Abstract: |
| Objective To compare the diagnostic efficacy of computer-aided diagnosis (CAD) system based on different guidelinesSfor the diagnosis of benign and malignant thyroid nodules. Methods This multicentric prospective study analyzed 453 patients with 521 thyroid nodules visiting the four hospitals of federation between January 2019 and September 2019. All the nodules were confirmed by pathology of surgery or ultrasound guided fine needle aspiration . Using pathology as the gold standard, establish the receiver operating characteristic curve (ROC curve), and obtain the area under the curve (AUC), cut-off value and its corresponding sensitivity, specificity, positive predictive value (PPV), negative Predicted value (NPV). The Z-test was used to compare the AUC between groups. The McNemar test was used to compare the sensitivity and specificity between groups. Result The AUC and sensitivity of the KSThR consensus is higher than the other seven guidelines (P<0.01), the specificity of the BTA guidelines was higher than the other seven guidelines (P<0.01). Conclusion The KSThR consensus has the highest AUC, sensitivity, and the BTA guidelines have the highest specificity. |
| Key words: Thyroid nodule Ultrasonography Computer-aided system |
