基于超声的预测模型用于区分伴周边钙化的甲状腺结节良恶性
研究纳入380个伴周边钙化的甲状腺结节,按268个训练队列和112个验证队列开发并验证超声列线图模型。晕征、钙化外突、周边钙化类型、边缘、内部回声和组成是恶性独立预测因子,训练与验证队列AUC分别为0.904和0.882。模型可能有助于预测恶性风险并提高诊断准确性。
要点
- 伴周边钙化的甲状腺结节中,晕征、钙化外突、周边钙化类型、边缘、内部回声和组成是恶性独立预测因子。
- 基于上述6个变量的超声列线图在训练队列AUC为0.904,验证队列AUC为0.882。
- 决策曲线分析显示概率阈值0.11-0.83可使患者临床获益。
核心结论
- 基于超声的列线图模型可能预测伴周边钙化甲状腺结节的恶性风险,从而帮助提高诊断准确性。
研究设计与样本
- 回顾性纳入2017年1月至2025年1月超声检查的8359个甲状腺结节中380个伴周边钙化且有病理结果和超声检查的结节,268个进入训练队列、112个进入验证队列;候选变量包括年龄、性别和灰度超声图像特征;通过单因素和多因素分析确定恶性独立危险因素并建立列线图模型,用AUC、校准曲线和决策曲线分析评估性能。
关键影像表现/征象
- 晕征、钙化外突、周边钙化类型、边缘、内部回声和组成是伴周边钙化甲状腺结节恶性的显著独立预测因子。
诊断标准与关键数值
- 基于6个变量的列线图模型在训练队列AUC为0.904,在验证队列AUC为0.882。
- 决策曲线分析显示概率阈值0.11-0.83可使患者临床获益。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
RATIONALE AND OBJECTIVES: The differential diagnosis of thyroid nodules with peripheral calcifications by ultrasound (US) has always been source of confusion. This study aimed to develop and validate a predictive US-based nomogram model to differentiate malignant from benign nodules with peripheral calcifications.
METHODS: Of the 8359 thyroid nodules scanned by ultrasonography between January 2017 and January 2025, 380 nodules with peripheral calcifications were included, with confirmed pathological results and US examinations. 268 nodules were included in the training cohort, and 112 nodules were included in the validation cohort. The candidate variables included age, gender, and the image features obtained from grayscale US. Independent risk factors for malignant thyroid nodules were determined by univariate and multivariate analyses, and a predictive nomogram model was developed. The performance of the US nomogram was assessed by the area under the curve (AUC), calibration curve, and decision curve analysis (DCA) results.
RESULTS: Univariate and multivariate logistic regression analyses revealed that the halo sign, extrusion beyond calcification, type of peripheral calcification, margin, internal echogenicity, and composition were significant independent predictors for malignant thyroid nodules with peripheral calcifications. The nomogram model based on the six variables exhibited excellent calibration and discrimination in the training and validation cohorts, with AUC values of 0.904 and 0.882, respectively. The DCA showed that a probability threshold of 0.11-0.83 could benefit patients clinically.
CONCLUSION: The US-based nomogram model can potentially predict the malignant risk of thyroid nodules with peripheral calcifications, thereby helping to enhance diagnostic accuracy.
引用
- Bai S, Wu L, Su Y, et al. Ultrasound-Based Prediction Model for Distinguishing Malignant from Benign Thyroid Nodules with Peripheral Calcification. Acad Radiol. 2025. PMID: 40393828. doi:10.1016/j.acra.2025.04.073
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。