文献速递乳腺MR

超快速DCE-MRI动力学参数鉴别乳腺良恶性病变

Value of kinetic parameters of ultra-fast dynamic contrast-enhanced magnetic resonance inaging (MRI) for differentiating malignant and benign breast lesions: a retrospective study.
PubMed 文献 · Clin Radiol · 2026-08-14

回顾性纳入194例乳腺病变患者(94例恶性、100例良性),分为训练集139例与测试集55例,均行超快速动态增强MRI。训练集中最大斜率、时间到强化、曲线下面积、最大峰值在良恶性间均有显著差异,但时间到强化诊断效能有限。最大斜率、曲线下面积、最大峰值在肿块样强化中鉴别价值较高,在非肿块样强化中无显著差异。

要点

  • 超快速DCE-MRI的最大斜率、曲线下面积、最大峰值对乳腺良恶性鉴别有较好效能。
  • 时间到强化虽有统计学差异,但诊断效能有限,不宜单独用于定性。
  • 上述参数在肿块样强化中价值较高,对非肿块样强化鉴别意义不大。

核心结论

  • 超快速DCE-MRI动力学参数有助于鉴别乳腺良恶性病变,对肿块样强化病变效能较好。
  • 最大斜率、曲线下面积、最大峰值具有较好的鉴别能力,时间到强化诊断效能有限。
  • 在非肿块样强化亚组中,各参数在良恶性病变间均无显著差异。

研究设计与样本

  • 单中心回顾性研究,纳入194例乳腺病变患者(194个病灶),其中恶性94例、良性100例。
  • 纳入时间为2024年2月至2025年6月,按训练集139例与测试集55例划分。
  • 所有患者均接受超快速动态对比增强磁共振成像检查。
  • 训练集用于比较良恶性病变间参数差异并确定阈值,测试集用于验证分类准确性。

关键影像表现/征象

  • 最大斜率、曲线下面积、最大峰值、时间到强化在良恶性乳腺病变间均显示显著差异。
  • 在肿块样强化亚组中,最大斜率、曲线下面积、最大峰值在良恶性间仍保持显著差异。
  • 在非肿块样强化亚组中,最大斜率、曲线下面积、最大峰值在良恶性间未见显著差异。

诊断标准与关键数值

  • 训练集中最大斜率、曲线下面积、最大峰值的ROC-AUC分别为0.80、0.82、0.79。
  • 时间到强化的ROC-AUC为0.37,诊断效能有限。
  • 测试集中最大斜率、曲线下面积、最大峰值的ROC-AUC分别为0.80、0.83、0.80。
  • 测试集中最大斜率、曲线下面积、最大峰值的敏感度分别为0.84、0.89、0.84。
  • 测试集中最大斜率、曲线下面积、最大峰值的特异度分别为0.78、0.78、0.86。

临床应用与阅片要点

  • 阅片时优先参考最大斜率、曲线下面积、最大峰值,对肿块样强化病灶的定性价值较高。
  • 时间到强化单独使用诊断价值有限,不建议作为定性依据。
  • 对非肿块样强化病变,上述参数鉴别良恶性意义有限,应结合形态学及其他序列综合判断。

局限性

  • 本研究为回顾性设计,可能产生选择偏倚。
  • 研究基于单中心数据,未提及外部验证。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

AIM: The aim of this study was to evaluate the diagnostic value of ultra-fast dynamic contrast-enhanced magnetic resonance imaging (UF DCE-MRI) kinetic parameters in differentiating benign and malignant breast lesions.

MATERIALS AND METHODS: A total of 194 patients (194 lesions: 94 malignant and 100 benign) were retrospectively recruited between February 2024 and June 2025 and divided into a training set (n = 139) and a test set (n = 55). All patients underwent UF DCE-MRI. In the training set, differences in maximum slope (MS), time to enhancement (TTE), area under the curve (AUC), and maximum peak of the curve (MPC) between benign and malignant lesions were compared. Furthermore, these parameters were analysed in both mass enhancement (ME) and non-mass enhancement (NME) subgroups. Subsequently, the optimal cut-offs derived from the training set were applied to the test set to calculate the classification accuracy of each parameter. A receiver operating characteristic (ROC) curve analysis was performed, with a P value <0.05 considered statistically significant.

RESULTS: All four parameters differed significantly between benign and malignant lesions (all P < .05). Notably, while TTE exhibited statistical significance; its diagnostic efficacy was limited (area under the receiver operating characteristic curve [ROC-AUC]: 0.37), whereas MS, AUC, and MPC achieved favourable discriminative ability (ROC-AUC: 0.80, 0.82, 0.79). In subgroup analysis, these three parameters remained significantly different between benign and malignant breast lesions in ME (all P < .05), but no significant differences were observed in NME (all P > .05). Consistently, they demonstrated high diagnostic efficacy in the test set (ROC-AUC: 0.80, 0.83, 0.80; sensitivity: 0.84, 0.89, 0.84; specificity: 0.78, 0.78, 0.86).

CONCLUSION: UF DCE-MRI kinetic parameters are promising for differentiating benign and malignant breast lesions, with particular efficacy in ME lesions.

来源:Clin Radiol,查看原文。版权归原作者所有。

引用

  1. Li, Wang, Nie, et al. Value of kinetic parameters of ultra-fast dynamic contrast-enhanced magnetic resonance inaging (MRI) for differentiating malignant and benign breast lesions: a retrospective study. Clin Radiol. 2026. PMID: 42727533. doi:10.1016/j.crad.2026.107459

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。