摘要
目的:探讨超快速动态对比增强磁共振(UF DCE-MRI)预测乳腺癌患者乳头乳晕复合体(NAC)受侵犯的价值。方法:回顾性分析于本院进行UF DCE-MRI检查的乳腺癌患者,对患者临床病理信息、UF DCE-MRI影像征象及动力学参数进行组间差异分析,使用Logistic回归构建乳腺癌侵犯NAC预测模型,采用受试者工作特征(ROC)曲线评估模型的诊断效能。结果:NAC受侵犯组肿瘤最大直径大于未侵犯组(P<0.001);NAC受侵犯组早期强化比例、腋窝淋巴结转移率显著高于未侵犯组,无强化率低于未侵犯组(P均<0.001);NAC受侵犯组的Ktrans、Ve显著低于未侵犯组,而kep、MS则显著高于未侵犯组(P均<0.05)。多因素Logistic回归分析筛选出NAC强化模式、Ktrans、kep、MS是乳腺癌侵犯NAC的独立预测因子,构建的乳腺癌侵犯NAC预测模型的曲线下面积(AUC)为0.89,最佳截断值为0.58,敏感度及特异度分别为82.91%、81.82%,其中Ktrans贡献度最高(32.4%)。结论:基于UF DCE-MRI影像特征构建的乳腺癌侵犯NAC预测模型具有较高的效能。
Objective:To explore the predictive value of ultrafast dynamic contrast-enhanced magnetic resonance imaging(UF DCE-MRI)in evaluating the invasion of nipple areola complex(NAC)in breast cancer patients.Methods:A retrospective analysis was conducted on the breast cancer patients examined by UF DCE-MRI in our hospital.The patient's clinical pathological information,imaging features and the dynamic parameters were analyzed between groups.Logistic regression was used to construct a prediction model for NAC invasion in breast cancer.The diagnostic efficacy of the model was evaluated using receiver operating characteristic curves(ROC).Results:The maximum tumor diameter was significantly larger in the NAC-invaded group than in the non-invaded group(P<0.001).The rates of NAC early enhancement and axillary lymph node metastasis were significantly higher,while the rate of non-enhancement was significantly lower,in the NAC-invaded group compared to the non-invaded group(P<0.001).The K^(tras) and V_(e) values in the invaded group were significantly lower than those in the non invaded group,and the kep and MS values in the invaded group were significantly higher than those in the non invaded group(P<0.05).Multivariate Logistic regression analysis showed that NAC enhancement pattern,K^(trans),k_(ep) and MS were independent predictors of NAC invasion,the area under the ROC curve(AUC)was 0.89.The sensitivity and specificity were 82.91%and 81.82%,respectively the cut-off value was 0.58.K^(trans) had the highest contribution(SHAP value=32.4%).Conclusion:The predictive model for NAC invasion in breast cancer based on UF DCE-MRI has high diagnostic efficacy.
作者
王梅
刘斐
沈世田
郭浩东
李海歌
WANG Mei;LIU Fei;SHEN Shi-tian;GUO Hao-dong;LI Hai-ge(Department of Medical Imaging,the Second Affiliated Hospital of Nanjing Medical University,Nanjing 210011,China)
出处
《中国临床医学影像杂志》
北大核心
2026年第3期178-181,195,共5页
Journal of China Clinic Medical Imaging
基金
南京医科大学科技发展基金项目(NMUB20240019)。
关键词
乳腺肿瘤
磁共振成像
Breast Neoplasms
Magnetic Resonance Imaging
