| 摘要: |
| 摘 要 目的 本研究旨在探讨剪切波弹性成像（SWE）联合常规超声多参数权重法新模型在甲状腺结节良恶性鉴别诊断中的价值。方法 回顾性分析2021年6月到2022年9月具有明确病理结果和完整常规超声及SWE资料的168例患者的175个甲状腺结节，其中138例连续患者的145个结节被纳入建模队列，30例连续患者的30个结节被纳入验证队列。本研究中，所有甲状腺结节均经术后病理证实或超声引导下细针穿刺抽吸细胞学检查（US-FNAC）获得明确诊断。采用单因素分析确定常规超声特征和SWE参数的危险因素；采用二元logistic回归分析筛选独立危险因素，计算每个危险因素的比值比（OR）；根据相关指标的OR值赋分，计算每个结节总分值，绘制受试者工作特性曲线（ROC），建立基于多参数权重的预测模型，计算曲线下面积（AUC）来评估模型的诊断性能；并在验证队列中进行验证。结果 SWE参数中ROC显示最大弹性模量的AUC（0.813）高于平均弹性模量（0.730），最佳临界点为38.3kPa，是区分甲状腺良恶性结节的最佳参数（P＜0.001），常规超声参数中纵横比≥1、实性成分、点状强回声是独立的危险因素；根据不同参数的OR值获得基于多参数权重的预测模型，建模队列鉴别甲状腺结节良恶性的AUC=0.906，95%可信区间0.850-0.961，灵敏度87.4%，特异性90.5%，PPV95.7%，NPV74.5%，准确性88.0%；验证队列鉴别甲状腺结节良恶性的AUC=0.918，95%可信区间0.808-0.985，灵敏度为91.9%，特异性75.0%，PPV 91.9%，NPV 75.0%，准确性87.0%。结论 SWE联合常规超声多参数权重法新模型有助于提升超声多模态成像技术在甲状腺良恶性结节中的应用价值，为临床实践和US-FNAC提供简便、经济的选择策略。 |
| 关键词: 甲状腺结节 剪切波弹性成像 超声特征 风险分层 多参数权重法 |
| DOI： |
| 投稿时间：2022-11-10修订日期：2023-01-26 |
| 基金项目:中国超声医师科技新星计划课题(KJXX2021001) |
|
| SWE combined with conventional ultrasound multi-parameter weight method to distinguish benign and malignant thyroid nodules |
| LI Zifeng,YAN Jiping,WANG Jinping,ZHU Lin,ZHANG Changfeng,LIU Sha |
| (Shanxi Provincial People’s Hospital) |
| Abstract: |
| ABSTRACT Objective The purpose of this study was to investigate the value of a new model of shear wave elastography (SWE) combined with conventional ultrasound multi-parameter weighting method in the differential diagnosis of benign and malignant thyroid nodules. Methods A retrospective analysis was performed on 175 thyroid nodules in 168 patients with definite pathological findings and complete conventional ultrasound and SWE data from June 2021 to September 2022, among which 145 nodules in 138 consecutive patients were included in the modeling cohort and 30 nodules in 30 consecutive patients were included in the validation cohort. In this study, all thyroid nodules were confirmed by postoperative pathology or ultrasound-guided fine-needle aspiration cytology (US-FNAC). Single factor analysis was used to determine the risk factors of conventional ultrasound features and SWE parameters. Binary logistic regression analysis was used to screen independent risk factors, and the odds ratio (OR) of each risk factor was calculated. According to the OR values of relevant indicators, the total score of each nodule was calculated, the receiver operating characteristic curve (ROC) was drawn, the prediction model based on the weight of multiple parameters was established, and the area under the curve (AUC) was calculated to evaluate the diagnostic performance of the model. And validate in the validation queue. Results In SWE parameters, ROC showed that the AUC of maximum elastic modulus (0.813) was higher than the average elastic modulus (0.730), and the optimal critical point was 38.3kPa, which was the best parameter to distinguish thyroid benign and malignant nodules (P < 0.001). In conventional ultrasound parameters, aspect ratio≥1, solid component and echogenic foci were independent risk factors. According to the OR values of different parameters, the prediction model based on the weight of multiple parameters was obtained. The AUC of the model was 0.906, the 95% confidence interval was 0.850-0.961, the sensitivity was 87.4%, the specificity was 90.5%, the PPV was 95.7%, the NPV was 74.5%, and the accuracy was 88.0%. The AUC for the differentiation of benign and malignant thyroid nodules in the validation cohort was 0.918, 95% confidence interval 0.808-0.985, sensitivity 91.9%, specificity 75.0%, PPV 91.9%, NPV 75.0%, accuracy 87.0%. Conclusion The new model of SWE combined with conventional ultrasound multi-parameter weighting method is helpful to improve the application value of ultrasonic multimodal imaging technology in benign and malignant thyroid nodules, and provide a simple and economical selection strategy for clinical practice including US-FNAC. |
| Key words: Thyroid nodule Shear wave elastic imaging Ultrasonic characteristics Stratification of risk Multi-parameter weight method |
