初步探讨建立CT检查中甲状腺偶发结节的分类标准  

A New Thyroid Imaging Reporting and Data System to Categorize Incidental Thyroid Nodules Found in CT Scan

作　　者：陶全 陈克敏[1] 周永明[2] 谭文莉[3] Jie-ming FANG 何文辉 罗禹 孙建[5] 祝颖 赵俊松 王明明[2] 马为彬 TAO Quan;CHEN Ke-min;ZHOU Yong-ming;TAN Wen-li;Jie-ming FANG;HE Wen-hui;LUO Yu;SUN Jian;ZHU Ying;ZHAO Jun-song;WANG Ming-ming;MA Wei-bin(Department of Radiology,Ruijin Hospital,Shanghai Jiaotong University School of Medicine;Department of Radiology,Shanghai Fourth People's Hospital,Tongji University;Department of Radiology,Shuguang Hospital,Shanghai University of TCM;Department of Interventional Radiology,City of Hope National Medical Center,Duarte,CA,United States;Department of Surgery,Shanghai Fourth People's Hospital,Tongji University;Department of Pathology,Shanghai Fourth People's Hospital,Tongji University;Department of Radiology,Shanghai Baoshan Hospital of Integrated Traditional Chinese and Western Medicine)

机构地区：[1]上海交通大学医学院附属瑞金医院放射科 [2]同济大学附属上海市第四人民医院放射科 [3]上海中医药大学附属曙光医院放射科 [4]美国南加利福利亚州希望之城医院介入放射科 [5]同济大学附属上海市第四人民医院外科 [6]同济大学附属上海市第四人民医院病理科 [7]上海市宝山区中西医结合医院放射科

出　　处：《中国医学计算机成像杂志》2019年第6期520-525,共6页Chinese Computed Medical Imaging

基　　金：上海市第四人民医院科技创新基金项目资助(编号：2016SYFKJ11)~~

摘　　要：目的:针对CT检査中甲状腺偶发结节的分类问题,探讨建立基于CT征象的甲状腺影像报告和数据系统(CT-TIRADS)o方法:收集我院甲状腺结节手术证实病例152例,183个结节。与病理结果对照,筛选出CT恶性结节的征象,参照Kwak超声的TIRADS制定CT-TIRADS,计算各类结节的实际恶性率。通过ROC曲线下面积比较各CT征象与CT-TIRADS诊断甲状腺结节良恶性的诊断效能。结果:甲状腺结节的6个CT恶性征象包括:非纯囊性、单发、钙化、边缘不规则、周围浸润、淋巴结转移。CT-TIRADS1-3类、4a-4c类及5类结节恶性率分别为:0%、0%,1.4%、12.2%、56.0%,75%,100%。CT一TIRADS诊断甲状腺结节良恶性的ROC曲线下面积为0.919,明显高于其他6个恶性CT征象。结论:本文建立了CTTIRADS,其对CT检査的甲状腺结节分类具有较高的临床价值。Purpose:To develop a new thyroid imaging reporting and data system based on CT features(CT-TIRADS),which could be used to categorize incidental thyroid nodules found in CT scan.Methods:We included 183 thyroid nodules in 152 patients with neck CT data and thyroidectomy results.According to Kawk's TIRADS,we designed a new TIRADS based on CT features named CT-TIRADS.ROC curves were performed to compare the diagnostic value of CT-TIRADS.Results:The features of non-cystic,single nodule,irregular margins,calcification,surrounding invasion and lymphatic metastasis were related to the malignant thyroid nodules.Malignant risks of category 1 to 3,4a to 4c,and 5 were 0%,0%,1.4%,12.2%,56.0%,75%,100%,according our data.The area under the ROC curve of CT-TIRADS was 0.919,lager than other 6 CT features.Conclusion:The new TIRADS based on CT features could be used to accurately categorize the thyroid nodules found in CT scan.It would be useful to manage incidental thyroid nodules.

关 键 词：甲状腺偶发结节 体层摄影术 X线计算机 甲状腺影像报告和数据系统 

分 类 号：R445.3[医药卫生—影像医学与核医学]