基于多模态超声的列线图模型鉴别BI-RADS 4类乳腺病变的临床价值
Clinical value of nomogram model based on multimodal ultrasound in differentiating benign and malignant lesions in BI-RADS 4 category breast lesions
周玮珺 1刘勇 1徐平 1吴兰英 1王颖 2肖榕 3杨敏1
作者信息
- 1. 100038 北京市,首都医科大学附属北京世纪坛医院超声科
- 2. 南京鼓楼医院超声科
- 3. 安徽医科大学第一附属医院超声科
- 折叠
摘要
目的 构建基于临床资料、剪切波弹性成像参数和超声影像组学的列线图模型,探讨其鉴别BI-RADS 4类乳腺病变良恶性的临床价值.方法 回顾性收集3家医院共403例BI-RADS 4类乳腺病变患者(共403个病灶)的临床资料、剪切波弹性成像及病理检查资料,以2017年12月至2019年6月南京鼓楼医院和2019年6～12月安徽医科大学第一附属医院共282个病灶为训练集,2022年4月至2023年6月北京世纪坛医院的121个病灶为验证集,根据病理结果将训练集和验证集分别分为良性组和恶性组.通过提取病灶灰阶超声影像组学特征计算影像组学评分(Rad-score).采用单因素及多因素Logistic回归分析筛选鉴别BI-RADS 4类乳腺病变良恶性的影响因素,构建预测模型并绘制列线图;采用受试者工作特征曲线、校准曲线及临床决策曲线评估该模型的诊断效能.结果 训练集及验证集中恶性组年龄、病灶大小,以及病灶剪切波速度最大值、最小值、平均值和中位数(SWVmax、SWVmin、SWVmean和SWVmedian)均高于良性组,差异均有统计学意义(均P<0.001).经过特征提取及筛选,最终纳入13个影像组学特征用于计算Rad-score,验证集良、恶性组Rad-score分别为-1.07(-1.64,-0.37)分、0.07(-0.30,0.56)分,二者比较差异有统计学意义(P<0.001).多因素Logistic回归分析显示,年龄、SWVmax及Rad-score均为鉴别乳腺病变良恶性的独立影响因素(OR=1.107、3.919、4.180,均P<0.001).基于以上3个因素联合应用构建的列线图模型在训练集及验证集中鉴别BI-RADS 4类乳腺病变良恶性的曲线下面积均高于SWVmax和Rad-score单独应用(均P<0.05);校准曲线显示该模型的校准度均高(均P>0.05);在验证集中使用列线图模型鉴别BI-RADS 4类乳腺病变良恶性能获得更高的临床收益,将非必要穿刺活检率降低了61.16%.结论 基于患者年龄、SWVmax及Rad-score构建的列线图模型能有效鉴别BI-RADS 4类乳腺病变良恶性,降低非必要穿刺活检率,有一定的临床价值.
Abstract
Objective To establish a nomogram model based on clinical data,shear wave elastography(SWE)parameters and ultrasound imaging radiomics,and to explore the clinical value of the model in differentiating benign and malignant BI-RADS 4 category lesions.Methods The clinical data,shear wave elastography and pathological examination results of 403 patients with BI-RADS 4 category lesions from 3 hospitals were retrospectively collected.A total of 282 breast lesions in Nanjing Drum Tower Hospital from December 2017 to June 2019 and the First Affiliated Hospital of Anhui Medical University from June to December 2019 were selected as training set.A total of 121 breast lesions in Beijing Shijitan Hospital from April 2022 to June 2023 were selected as validation set.According to pathological results,the training set and the validation set were divided into benign group and malignant group.The radiomics features of B-mode ultrasound of the lesions were extracted and the radiomics score(Rad-score)was calculated.Univariate and multivariate Logistic regression analysis were used to screen the influencing factors for differentiating benign and malignant BI-RADS 4 category lesions,a prediction model was constructed and a nomogram was drawn.Receiver operating characteristic(ROC)curve,calibration curve and clinical decision curve were used to evaluate the diagnostic efficacy of the model.Results In the training set and validation set,the age,lesion size,and maximum,minimum,mean and median values of SWE(SWVmax,SWVmin,SWVmean and SWVmedian)in the malignant group were higher than those in the benign group,and the differences were statistically significant(all P<0.001).After feature extraction and screening,13 radiomics features were finally included in the calculation of Rad-score.The Rad-score of the benign and malignant groups in the validation set were-1.07(-1.64,-0.37)scores and 0.07(-0.30,0.56)scores,respectively,the difference was statistically significant(P<0.001).Multivariate Logistic regression analysis showed that age,SWVmax and Rad-score were independent influencing factors in predicting benign and malignant breast lesions(OR=1.107,3.919,4.180,all P<0.001).The nomogram model was established based on above three indexes.The area under the curve of the nomogram model was higher than that of SWVmax and Rad-score in the training set and validation set(both P<0.05),and the calibration curve showed the fitting degree were good(both P>0.05).In the validation set,the nomogram model could achieve higher clinical benefits in predicting benign and malignant BI-RADS 4 category lesions and could reduce the unnecessary biopsy rate by 61.16%.Conclusion The nomogram model based on age,SWVmax and Rad-score can effectively differentiating benign and malignant BI-RADS 4 category lesions,and reduce the unnecessary biopsy rate,which has certain clinical value.
关键词
超声检查/剪切波弹性成像/影像组学/列线图/BI-RADS/4类/乳腺病变,良恶性Key words
Ultraonography/Shear wave elastography/Radiomics/Nomogram/BI-RADS 4 category/Breast lesions,benign and malignant引用本文复制引用
出版年
2024
NETL
NSTL
维普
万方数据