标题：MRI动态增强定量参数对乳腺非肿块样强化良恶性病变的鉴别诊断价值

      作者：李晏，郑晓林，刘碧华，邹玉坚    东莞市人民医院放射科，广东 东莞 523000

      卷次： 2020年31卷10期

      【摘要】 目的 探讨动态增强磁共振成像(DCE-MRI)定量参数对乳腺非肿块样强化(NME)良恶性病变的鉴别诊断价值。方法 选择2014年6月至2015年9月间在东莞市人民医院行DCE-MRI扫描的84例NME患者作为研究对象，所有患者的增强后表现均为乳腺非肿块样强化并经病理学诊断明确其病变性质。根据最终病理学诊断结果分为良性病变组30例和恶性病变组54例，良性病变组又分为乳腺增生组16例和乳腺炎症组14例，恶性病变组又分为导管原位癌组(DCIS，23例)和浸润性导管癌组(IDC，31例)。对比分析良性病变组、恶性病变组以及两组中各亚组的容量转移常数(Ktrans)、速率常数(Kep)及血管外细胞外间隙容积比(Ve)等定量参数。以病理结果为金标准，绘制受试者工作特征曲线(ROC)评价各定量参数在诊断NME良恶性病变的诊断效能。结果 良性病变组患者的Ktrans、Kep、Ve均值分别为(0.070±0.014)/min、(0.490±0.079)/min、0.173±0.034，恶性病变组患者分别为(0.111±0.025)/min、(0.767±0.106)/min、0.168±0.029，两组间Ktrans、Kep比较差异均有统计学意义(P<0.05)，但两组间的Ve比较差异无统计学意义(P>0.05)。DCIS组及 IDC组患者的 Ktrans、Kep、Ve均值分别为(0.087±0.026)/min、(0.665±0.187)/min、0.142±0.035和(0.147±0.041)/min，(0.838±0.214)/min、(0.152±0.041)，两亚组间Ktrans、Kep比较差异均有统计学意义(P<0.05)，但Ve比较差异无统计学意义(P>0.05)；乳腺增生组、乳腺炎症组患者的Ktrans、Kep、Ve均值分别为(0.055±0.012)/min、(0.475±0.089)/min、0.179±0.046和(0.107±0.023)/min、(0.503±0.095)/min、0.166±0.038，两亚组间的Ktrans比较差异有统计学意义(P<0.05)，但Kep、Ve比较差异均无统计学意义(P>0.05)；以最大约登指数为最佳诊断切入点，Ktrans、Kep、Ve在诊断非肿块强化恶性病变上的AUC分别为 0.741、0.767和 0.434，三者的敏感性分别为 81.48%、79.63%、53.70%，特异性分别为 76.67%、83.33%、53.33%，准确性分别为77.38%、80.95%、53.57%，Ktrans、Kep的诊断效能比Ve更好，三者在敏感性、特异性和准确性方面比较差异有统计学意义(P<0.05)。结论 DCE-MRI定量参数Ktrans、Kep在鉴别诊断NME良恶性病变上具有较高的诊断效能，可作为NME病变定性诊断的可靠指标应用于临床，有望成为无创评价乳腺肿瘤微循环的新方法。
      【关键词】 动态增强磁共振成像；定量参数；乳腺；非肿块强化病变；良恶性；鉴别诊断
      【中图分类号】 R445 【文献标识码】 A 【文章编号】 1003—6350（2020）10—1283—05

Differential diagnosis of non-mass-like enhancement benign and malignant breast lesions with MRI dynamiccontrast-enhanced quantitative parameters.
LI Yan, ZHENG Xiao-lin, LIU Bi-hua, ZOU Yu-jian. Department ofRadiology, Dongguan People's Hospital, Dongguan 523000, Guangdong, CHINA
【Abstract】 Objective To research the diagnostic value of quantitative parameters of dynamic contrast-en-hanced magnetic resonance imaging (DCE-MRI) in differentiating non-mass-like enhancement (NME) benign and malig-nant breast lesions. Methods From June 2014 to September 2015, 84 NME patients who underwent DCE-MRI scan inDongguan People's Hospital were selected as the study objects. All patients' enhanced manifestations were non-mass-likeenhancement of breast, and their pathological nature were confirmed by pathological diagnosis. According to the finalpathological diagnosis, they were divided into benign lesion group (30 cases) and malignant lesion group (54 cases). Thebenign lesion group was divided into two subgroups: hyperplasia group (16 cases) and inflammation group (14 cases).The malignant lesion group was divided into ductal carcinoma in situ (DCIS, 23 cases) group and invasive ductal carcino-ma (IDC, 31 cases). Quantitative parameters such as volume transfer constant (Ktrans), rate constant (Kep) and extracellularspace volume ratio (Ve) were compared and analyzed in benign lesion group, malignant lesion group and each subgroup ofthe two groups. The pathological results were taken as the gold standard, and the receiver operating characteristic curve(ROC) was drawn to evaluate the diagnostic efficiency of each quantitative parameter in diagnosing NME benign and ma-lignant lesions. Results The mean values of Ktrans , Kep and Ve in benign lesion group were (0.070±0.014)/min, (0.490±0.079)/min, 0.173±0.034, respectively, versus corresponding (0.111±0.025)/min, (0.767±0.106)/min, 0.168±0.029 in ma-lignant lesion group. There were significant differences in Ktrans, Kep between the two groups (P<0.05), and no significantdifference in Ve between the two groups (P>0.05). The mean values of Ktrans, Kep and Ve in DCIS group and IDC groupwere (0.087±0.026)/min, (0.665±0.187)/min, 0.142±0.035 and (0.147±0.041)/min, (0.838±0.214)/min, 0.152±0.041, re-spectively. There were significant differences between Ktrans and Kep between the subgroups (P<0.05), and no significantdifference in Ve between the two groups (P>0.05). The mean values of Ktrans, Kep and Ve in breast hyperplasia groupand inflammation group were (0.055 ± 0.012)/min, (0.475 ± 0.089)/min, 0.179 ± 0.046 and (0.107 ± 0.023)/min, (0.503 ±0.095)/min, 0.166 ± 0.038, respec