摘要
目的:分析不同乳腺癌的多模态超声图像特征,探讨其在分子分型中的诊断价值。方法:纳入本院120例乳腺癌患者,以病理诊断为金标准,提取常规超声(ultrasonic imaging,US)、应变超声弹性成像(strain ultrasound elastography,SUE)及超声造影(contrast-enhanced ultrasound,CEUS)的检查结果及相关临床数据。采用卡方检验筛选组间差异特征,双向逐步Logistic回归分析构建分子分型(Luminal型与非Luminal型)预测模型;通过ROC曲线评估单模态及多模态联合诊断效能。结果:US检查显示,Luminal型(LA/LB)与淋巴结形态、增强后范围、内部回声显著相关,非Luminal型(HER2^(+)/TNBC)与病灶形态、钙化、淋巴结形态及边界显著相关;SUE结果显示,各分子亚型均与弹性评分显著相关;CEUS结果显示,Luminal与非Luminal分型均与有无缺损、增强后范围显著相关;构建的预测模型AUC分别为0.859(Luminal型内部分型)和0.947(非Luminal型鉴别)。效能对比结果显示,多模态超声联合模型较单一模型具更优的诊断效能。结论:多模态超声可有效区分乳腺癌分子分型,联合模型显著提升诊断效能,为临床精准分型提供影像学依据。
Objective To analyze the multimodal ultrasound features of different breast cancers(BC)and explore their diagnostic value in molecular subtyping.Methods This study included 120 BC patients admitted to our hospital,with pathological diagnosis as the gold standard.Data from ultrasonic imaging(US),strain ultrasound elastography(SUE),contrast-enhanced ultrasound(CEUS),and relevant clinical information were collected.Chi-square tests were used to screen for significantly different features between groups,and two-way stepwise logistic regression analysis was performed to construct prediction models for molecular subtypes(Luminal vs.non-Luminal).Receiver operating characteristic(ROC)curves were used to evaluate the diagnostic efficacy of single-modal and multimodal combined ultrasound techniques.Results US examination showed that Luminal subtypes(LA/LB)were significantly correlated with lymph node morphology,enhanced range,and internal echo,while non-Luminal subtypes(HER2^(+)/TNBC)were significantly associated with lesion morphology,calcification,lymph node morphology,and margin.SUE results indicated that all molecular subtypes were significantly correlated with elasticity scores.CEUS analysis revealed that both Luminal and non-Luminal subtypes were significantly associated with enhancement defects and enhanced range.The area under the ROC curve(AUC)of the constructed prediction models was 0.859 for Luminal subtype differentiation and 0.947 for non-Luminal subtype discrimination.Efficacy comparison showed that the multimodal ultrasound combined model exhibited superior diagnostic performance compared with single-modal models.Conclusion Multimodal ultrasound can effectively distinguish BC molecular subtypes,and the combined model significantly improves diagnostic efficacy,providing an imaging basis for clinical precise subtyping.
作者
康锦涓
黄玲
于鹏
邱晶晶
蒋宇阳
王金荣
KANG Jinjuan;HUANG Ling;YU Peng;QIU Jingjing;JIANG Yuyang;WANG Jinrong(Taixing People's Hospital,Taizhou 225400)
出处
《晓庄学院学报(医学版)》
2025年第3期126-133,共8页
Journal of Hunan Normal University(Medical Sciences)
基金
2024年度江苏省青年科技人才托举工程(JSTJ-2024-615)
关键词
乳腺癌
多模态超声
超声造影
分子分型
breast cancer
multimodal ultrasound
contrast-enhanced ultrasound
molecular subtypes
