基于中国版甲状腺影像报告与数据系统的甲状腺结节恶性风险预测模型  

C-TIRADS-based nomogram for malignant risk prediction of thyroid nodules

作　　者：杨粒芝 孙霄[1] 商蒙蒙 郭鲁[1] 时丹丹[1] 李杰[1] YANG Lizhi;SUN Xiao;SHANG Mengmeng;GUO Lu;SHI Dandan;LI Jie(Department of Ultrasound,Qilu Hospital of Shandong University,Jinan 250012,Shandong,China)

机构地区：[1]山东大学齐鲁医院超声科,山东济南250012

出　　处：《山东大学学报（医学版）》2022年第6期64-69,81,共7页Journal of Shandong University:Health Sciences

摘　　要：目的构建基于中国版甲状腺影像报告与数据系统(C-TIRADS)列线图模型预测甲状腺结节恶性风险。方法收集2020年1月至2021年6月就诊于山东大学齐鲁医院甲状腺外科388例甲状腺结节患者的临床及超声资料,将2020年1月至2020年12月的270例患者作为建模组,2021年1月至2021年6月的118例患者作为验证组,根据术后常规病理结果,建模组分为良性结节组(n=137例)和恶性结节组(n=133例),通过单因素分析和多因素Logistic回归分析筛选出甲状腺恶性结节的独立危险因素,构建甲状腺结节恶性风险预测列线图模型并评价其性能。结果多因素Logistic回归分析提示年龄、促甲状腺激素(TSH)、甲状腺球蛋白(TG)以及C-TIRADS分类是甲状腺恶性结节的独立危险因素。基于以上独立危险因素构建的列线图模型,其预测建模组甲状腺恶性结节的受试者工作特征(ROC)曲线下面积(AUC)为0.981(95%CI:0.967~0.996),验证组使用列线图模型预测甲状腺结节恶性风险的AUC为0.951(95%CI:0.909~0.992),表明列线图具有出色的预测性能,列线图内部验证的一致性指数(C-index)为0.981,表明列线图的区分度良好。建模组校准曲线接近理想曲线,提示列线图预测甲状腺结节的恶性风险与实际风险间具有较高一致性。结论年龄、TSH、TG和C-TIRADS分类是预测甲状腺结节恶性风险的独立危险因素;基于C-TIRADS分类联合临床及血清学特征建立的列线图模型能够实现甲状腺结节恶性风险的个体化精准预测。Objective To establish a nomogram based on Chinese Thyroid Imaging Reporting and Data System(C-TIRADS)to predict the malignant risk of thyroid nodules.Methods The clinical and ultrasonic data of 388 patients with thyroid nodules treated at the Department of Thyroid Surgery of Qilu Hospital of Shandong University during Jan.2020 and Jun.2021 were collected.A total of 210 patients during Jan 2020 and Dec.2020 were selected as the development group,and 118 patients during Jan.2021 and Jun.2021 were selected as the validation group.According to the postoperative routine pathological results,the patients in the developwent group were divided into benign nodules group(n=137)and malignant nodules group(n=133).The independent risk factors of malignant thyroid nodules screened by univariate analysis and multivariate Logistic regression analysis were included to establish the nomogram.The performance of the nomogram was evaluated with concordance index(C-index),receiver operating characteristic(ROC)curve and calibration curve.Results Multivariate Logistic regression analysis revealed that age,thyroid stimulating hormone(TSH),thyroglobulin(TG)and C-TIRADS classification were independent risk factors for malignant thyroid nodules.The area under the ROC curve(AUC)of the nomogram was 0.981(95%CI:0.967-0.996),and the AUC of the validation group was 0.951(95%CI:0.909-0.992),showing the nomogram had excellent predictive performance.The C-index for internal validation of the nomogram was 0.981,suggesting that the nomogram had good discriminability.The calibration curve was close to the ideal curve,indicating that there was a high consistency between the predicted risk and actual risk.Conclusion Age,TSH,TG and C-TIRADS classification are independent risk factors for predicting malignant risk of thyroid nodules.The developed nomogram can realize the individualized and accurate prediction of malignant risk of thyroid nodules.

关 键 词：中国版甲状腺影像报告与数据系统 甲状腺结节 列线图 恶性风险 

分 类 号：R581[医药卫生—内分泌] R445.1[医药卫生—内科学]