基于多模态超声特征构建的列线图模型可有效鉴别肿块型浆细胞性乳腺炎与浸润性导管癌OA
A nomogram model based on multimodal ultrasound features can effectively differentiate mass-type plasma cell mastitis from invasive ductal carcinoma
目的 探讨基于二维超声(US)、超微血流成像(UMA)、应变弹性成像(SE)及超声造影(CEUS)的多模态特征构建的联合模型在鉴别肿块型浆细胞性乳腺炎(PCM)与浸润性导管癌(IDC)中的价值.方法 回顾性分析2024年9月~2026年5月在内蒙古自治区人民医院经病理证实为肿块型PCM患者74例和IDC患者157例,均行US、UMA、SE及CEUS检查.基于单因素分析中P<0.05的变量,经多因素logistic回归筛选独立预测因子后构建联合模型,并绘制列线图.采用ROC曲线评估联合模型的诊断效能,采用DeLong检验比较联合模型与各单模态的AUC差异,Bootstrap重抽样(1000次)进行内部验证.结果 年龄、垂直位生长、后方衰减、钙化、SE高硬度、血管弯曲度(VT)升高、AT缩短及AS升高为鉴别IDC的独立预测因子(均P<0.05).联合模型的AUC为0.946(95%CI:0.918~0.973),高于US(0.811)、SE(0.708)、UMA(0.732)及CEUS(0.784)(DeLong检验,P<0.001).最佳截断值下敏感度为76.4%,特异度为94.6%.Bootstrap内部验证校正后AUC为0.951.校准曲线显示模型拟合良好(Hosmer-Lemeshow检验P=0.721).结论 基于多模态超声特征构建的联合模型可显著提高肿块型PCM与IDC的鉴别效能,为临床无创诊断及个体化决策提供量化支持.
Objective To investigate the diagnostic value of a combined model based on multimodal ultrasound features,including conventional ultrasound(US),ultramicroangiography(UMA),strain elastography(SE),and contrast-enhanced ultrasound(CEUS),for differentiating mass-type plasma cell mastitis(PCM)from invasive ductal carcinoma(IDC).Methods A retrospective analysis was performed on 231 patients,including 74 patients with mass-type PCM and 157 patients with IDC,pathologically confirmed at Inner Mongolia Autonomous Region People's Hospital from September 2024 to May 2026.All patients underwent US,UMA,SE,and CEUS examinations.Variables with P<0.05 in univariate analysis were entered into multivariate logistic regression analysis to identify independent predictors,and a combined model was constructed and presented as a nomogram.ROC curves were used to evaluate the diagnostic performance of the combined model,and the differences in AUCs between the combined model and individual modalities were compared using the DeLong test.Internal validation was performed using 1000 bootstrap resamples.Results Age,vertical orientation,posterior acoustic attenuation,calcifications,high stiffness on SE,increased vessel tortuosity(VT),shortened arrival time(AT),and increased ascending slope(AS)were identified as independent predictors of IDC(P<0.05).The combined model achieved an AUC of 0.946(95%CI:0.918~0.973),which was higher than those of US(0.811),SE(0.708),UMA(0.732),and CEUS(0.784)(DeLong test,P<0.001).At the optimal cutoff value,the sensitivity and specificity were 76.4%and 94.6%,respectively.The corrected AUC after bootstrap internal validation was 0.951.The calibration curve demonstrated good model fit(Hosmer-Lemeshow test,P=0.721).Conclusion The combined model based on multimodal ultrasound features significantly improves the diagnostic performance for differentiating mass-type PCM from IDC and provides quantitative support for noninvasive diagnosis and individualized clinical decision-making.
刘梦月;王芳
内蒙古科技大学包头医学院,内蒙古 包头 014040||内蒙古自治区人民医院超声医学科,内蒙古 呼和浩特 010017内蒙古自治区人民医院超声医学科,内蒙古 呼和浩特 010017
多模态超声浆细胞性乳腺炎浸润性导管癌预测模型鉴别诊断
multimodal ultrasoundplasma cell mastitisinvasive ductal carcinomapredictive modeldifferential diagnosis
《分子影像学杂志》 2026 (8)
1022-1029,8
内蒙古自治区科技计划项目(2020GG0086)内蒙古自治区人民医院院内基金项目(201807)
评论