【作者】 郑斌；

【导师】 韩若凌；

【作者基本信息】 河北医科大学 ， 影像医学与核医学， 2019， 硕士

【摘要】 目的:比较Kwak等2011年制订的甲状腺影像报告和数据系统(简称Kwak TI-RADS)、美国放射学院2017年颁布的甲状腺影像报告和数据系统(ACR TI-RADS)和美国甲状腺协会2015年颁布的甲状腺结节恶性风险分层指南(简称ATA指南)对甲状腺结节的诊断价值。方法:对457例甲状腺结节患者(共计504个结节)的术前超声图像及报告进行回顾性分析,并将结节分别按照三种分类标准进行分类。以手术病理结果为“金标准”,为三种分类结果构建受试者工作特性曲线(ROC),分析三种分类标准对甲状腺良恶性结节的诊断价值,比较其诊断效能。结果:504个甲状腺结节中,根据ATA指南分为5类,分别为良性、极低度可疑、低度可疑、中度可疑、高度可疑,恶性率分别为0、0、3.1%、8.7%、74.9%,按ACR TI-RADS分为5类,分别为TR1、TR2、TR3、TR4、TR5,恶性率分别为0,0,2.7%,25.3%,76.4%,依据Kwak TI-RADS分为6类,分别为2类、3类、4a类、4b类、4c类和5类,各类别恶性率分别为0、0、10.1%、13.5%、72.5%、83.3%。ATA指南诊断甲状腺结节良恶性的准确度、敏感度、特异度、阳性预测值(PPV)、阴性预测值(NPV)及曲线下面积(AUC)分别为77.0%、98.1%、61.7%、65.1%、97.8%、0.8697。ACR TI-RADS诊断甲状腺良恶性结节的准确度、敏感度、特异度、阳性预测值(PPV)、阴性预测值(NPV)及曲线下面积(AUC)分别为75.2%、99.1%、57.9%、63.1%、98.8%、0.8746。Kwak TI-RADS诊断甲状腺结节良恶性的准确度、敏感度、特异度、阳性预测值(PPV)、阴性预测值(NPV)及曲线下面积(AUC)分别为77.0%、96.2%、63.0%、65.4%、95.8%、0.8839。三者间准确度、敏感度、特异度、PPV、NPV及AUC均未见统计学差异(P>0.05)。ATA分类中17个结节无法进行归类,约占甲状腺结节总数的3.4%(17/504),其中良性结节10个,约占2.0%(10/504),恶性结节7个,约占1.4%(7/504),未分类结节的恶性率约41.2%(7/17)。结论:三种分类标准均能为甲状腺结节提供有效的恶性风险分层方法,三者均可获得较高的敏感度及阴性预测值(NPV)。但ATA标准中未分类结节具有较高的恶性风险。更多还原

【Abstract】 Objective: To compare the diagnostic performance of thyroid imaging reporting and data system(TI-RADS)on thyroid nodules published by American College of Radiology(ACR TI-RADS)at 2017,Kwak(Kwak TI-RADS)at 2011 and 2015 American Thyroid Association(ATA)management guidelines.Mothed: Preoperative ultrasound images and reports of 457 thyroid nodules(504 nodules)were review and analyzed.Nodules were classified according to ATA guidelines and two TI-RADS criteria,and receiver operating characteristic curves of the classifications were constructed with pathological results as the gold standard.To analyze the diagnostic value of ATA guideline and two TI-RADS for thyroid nodules and compare their diagnostic performance.Result:For the 504 thyroid nodules,malignant rate for benign,very low suspicion,low suspicion,intermediate suspicion,and high suspicion by ATA guidelines was 0,0,3.1%,8.7%,74.9%,malignant rate for TR1-TR5 category patients classified by ACR TI-RADS was 0,0,2.7%,25.3%,76.4%,respectively;and that for 2,3,4a,4b,4c,5 category classified by Kwak TI-RADS was 0,0,10.1%,13.5%,72.5%,83.3%,respectively.Accuracy,susceptibility,specificity,positive predictive value(PPV),negative predictive value(NPV)and area under curve(AUC)of benign and malignant thyroid nodules diagnosed by ATA guidelines were 77.0%,98.1%,61.7%,65.1%,97.8%,0.8697,respectively.Accuracy,susceptibility,specificity,positive predictive value(PPV),negative predictive value(NPV)and area under curve(AUC)of benign and malignant thyroid nodules diagnosed by ACR TI-RADS were 77.0%,96.2%,63.0%,65.4%,95.8%,0.8839,respectively.Accuracy,susceptibility,specificity,positive predictive value(PPV),negative predictive value(NPV)and area under curve(AUC)of benign and malignant thyroid nodules diagnosed by Kwak TI-RADS were 77.0%,96.2%,63.0%,65.4%,95.8%,0.8839,respectively.No statistical difference in accuracy,susceptibility,specificity,PPV,NPV or AUC was found in the three groups(P>0.05).17 nodules could not be classified according to ATA criteria,3.4% of total thyroid nodules(17/504),in which 10 nodules were benign and 7 nodules were malignant.The malignant rate of ‘unclassified’ nodules was 41.2%(7/17).Conclusion:Both ATA guidelines and two TI-RADS performed quite well in diagnosing benign and malignant thyroid nodules,and all of them can obtain high sensitivity and negative predictive value.‘Unclassified’ nodules by ATA guidelines was at relatively high risk of malignancy.更多还原