常规超声联合声触诊组织成像量化技术对乳腺导管原位癌的诊断价值  

Diagnostic Value of Conventional Ultrasound Combined with Virtual Touch Tissue Imaging Quantification for Ductal Carcinoma In Situ of the Breast

作　　者：檀双秀 薛海燕 张一丹 陈桥梁 孔文韬 姚静 王颖 Tan Shuangxiu;Xue Haiyan;Zhang Yidan;Chen Qiaoliang;Kong Wentao;Yao Jing;Wang Ying(Department of Ultrasound Medicine,Nanjing Drum Tower Hospital(i.e.the Affiliated Hospital of Medical School of Nanjing University),Nanjing 210008,Jiangsu,China;Department of Nuclear Medicine,Nanjing Drum Tower Hospital(i.e.the Affiliated Hospital of Medical School of Nanjing University),Nanjing 210008,Jiangsu,China)

机构地区：[1]南京鼓楼医院/南京大学医学院附属鼓楼医院超声科,南京210008 [2]南京鼓楼医院/南京大学医学院附属鼓楼医院核医学科,南京210008

出　　处：《肿瘤预防与治疗》2024年第7期592-599,共8页Journal of Cancer Control And Treatment

摘　　要：目的:基于常规超声和声触诊组织成像量化(virtual touch tissue imaging quantification,VTIQ)技术构建一个乳腺导管原位癌(ductal carcinoma in situ,DCIS)的Nomogram预测模型。方法:回顾性收集本院2018年9月至2022年10月收治的乳腺导管内病变共165例患者(166处病灶)的影像学及临床资料,其中DCIS组58例,良性组108例。应用VTIQ技术测量乳腺病变的剪切波速度(shear wave velocity,SWV)相关参数,对年龄、常规超声特征及SWV相关参数进行单因素和多因素分析,得到DCIS的独立预测因子,并构建Nomogram预测模型。运用受试者工作特征曲线、校准曲线及决策曲线分析(decision curve analysis,DCA)以评估模型的诊断效能、准确性及临床实用性,并应用Z检验对各参数及预测模型的曲线下面积(area under the curve,AUC)进行组间比较。结果:年龄、最大径、最小径、形态、边界、微钙化、血流、乳腺影像报告数据系统(breast imaging reporting and data system,BI-RADS)、Vmax、Vmin、Vmean、Vmedian、离散系数在DCIS组与良性组间均具有统计学差异(P<0.05)。年龄、微钙化及Vmax为DCIS的独立预测因子(OR=1.055,P=0.004;OR=4.105,P=0.002;OR=1.511,P<0.001)。以此构建的Nomogram预测模型诊断DCIS的AUC值为0.846,显著高于BI-RADS(AUC=0.742,Z=3.143,P=0.002)。校准曲线显示,预测模型对DCIS的预测概率和实际概率较为一致。DCA分析显示,预测模型较BI-RADS可获得更大的临床净获益,表明该模型的临床实用性更强。结论:本研究联合常规超声和VTIQ技术构建的Nomogram预测模型,对DCIS具有较高的诊断效能,可为临床治疗决策制定提供参考。Objective:To construct a nomogram prediction model for ductal carcinoma in situ(DCIS)of the breast based on conventional ultrasound and virtual touch tissue imaging quantification(VTIQ).Methods:A total of 165 patients(166 lesions)with intraductal breast lesions admitted to our hospital from September 2018 to October 2022 were retrospectively collected,including 58 cases of DCIS and 108 cases of benign lesions.The parameters relevant to shear wave velocity(SWV)in breast lesions were measured using VTIQ.Univariate and multivariate analyses were performed on age,conventional ultrasound features and SWV-related parameters to determine independent predictors of DCIS,and then a nomogram prediction model was established.The diagnostic efficacy,accuracy and clinical practicality of the model were evaluated using the receiver operating characteristic(ROC)curve,calibration curve and decision curve analysis(DCA),and the area under the curve(AUC)in pairs were compared by Z test.Results:There were statistically significant differences in age,maximum diameter,minimum diameter,morphology,boundary,microcalcification,blood flow,BI-RADS,Vmax,Vmin,Vmean,Vmedian and coefficient of variation between the DCIS group and the benign group(P<0.05).Age,microcalcification and Vmax were independent predictors of DCIS(OR=1.055,P=0.004;OR=4.105,P=0.002;OR=1.511,P<0.001).The AUC value of the nomogram prediction model in the diagnosis of DCIS was 0.846,which was significantly higher than that of BI-RADS(AUC=0.742,Z=3.143,P=0.002).The calibration curve showed that the predicted probability of DCIS by the model was consistent with the actual probability for DCIS.The DCA showed that the prediction model could achieve greater clinical net benefit than BI-RADS,indicating stronger clinical practicality.Conclusion:The nomogram prediction model constructed by combining conventional ultrasound and VTIQ has high diagnostic efficacy for DCIS and can provide evidence for clinical treatment decision-making.

关 键 词：乳腺导管原位癌 超声 弹性成像 剪切波速度 预测模型 

分 类 号：R737.9[医药卫生—肿瘤]