量化 DCE-MRI 技术对乳腺良恶性病变诊断价值分析

目的：探讨基于动态增强磁共振成像（dynamic contrast-enhanced MRI，DCE-MRI）及特定药物动力学模型所获得的量化参数对乳腺良恶性病变诊断价值。方法：收集我院2015年6月至2016年2月42例乳腺病变患者行 DCE-MRI 扫描，通过后处理软件测量如下量化参数，定量参数：容量转移常数（volume transfer constant， Ktrans ）、速率常数（rate constant，Kep ）、血浆分数（the plasma fraction，Vp ）；半定量参数：（1）增强后病灶达峰时间（time to peak，TTP）、（2）增强后病灶内造影剂最大浓度值（max concentration，MAX Conc）、（3）增强后时间信号曲线下面积（area under curve，AUC）、（4）增强后时间信号曲线最大斜率值（MAX Slope）。检查后1周内均经手术取得病理诊断，采用独立样本 t 检验或非参数检验比较良性组及恶性组间各量化参数的差异。绘制 ROC 曲线，并分析有价值量化参数对乳腺良、恶性病灶诊断价值。结果：恶性病灶组定量参数 Ktrans Max（t ＝2．228，P ＝0．033）、Ktrans Mean（t ＝4．092，P ＜0．001）、Kep Mean（Z ＝2．422，P ＝0．015）、Kep Max（t ＝2．240，P ＝0．031）及半定量参数 MAX Conc Max（t ＝3．256，P ＝0．002）、MAX Conc Mean（t ＝3．460，P ＝0．001）、AUC Max（t ＝2．250，P ＝0．034）、AUC Mean（t ＝2．861，P ＝0．007）、MAX Slope Max（t ＝2．478，P ＝0．018）、MAX Slope Mean（t ＝2．226，P ＝0．032）较良性病灶组值明显升高，而半定量参数 TTP Min（t ＝－5．145，P ＜0．001）、TTP Mean（t ＝－3．818，P ＜0．001）较良性病灶组明显缩短。经 ROC 曲线分析显示，Ktrans Mean、Kep Mean、TTP Min、MAX Conc Max、AUC Mean、MAX Slope Max 对乳腺良恶性病变具有诊断价值，根据最大约登指数计算最佳诊断阈值，分别以0．1137、0．0258、0．9065、0．1059、0．1640、0.1502为诊断阈值，其鉴别诊断乳腺病灶良恶性敏感度分别为73．33％、73．33％、91．67％、76．67％、80．00％、80.00％，特异性分别为100％、83．30％、70．00％、91．70％、66．70％、75．00％。结论：基于 DCE-MRI 技术及特定的药物动力学模型所获得的量化参数对乳腺良恶性病变鉴别诊断有临床价值，且有较高的诊断效能，有望成为无创性评价乳腺肿瘤微循环的新方法。

The Diagnostic Features of Quantitative and Semi-quantitative Parameters Obtained from Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI)in Human Breast Lesions

Objective:To evaluate the quantitative and semi-quantitative parameters of Dynamic Contrast En-hanced MRI (DCE-MRI)based on a certain pharmacokinetic modeling in the diagnosis of human breast lesions.Methods:Forty-two patients with breast lesions were enrolled from June 2015 to Feb.2016.All patients received DCE-MRI examina-tion and were diagnosed pathologically within one week.The following quantitative parameters:volume transfer constant (Ktrans ),rate constant (Kep ),plasma fraction (Vp ),and semi-quantitative parameters:time to peak (TTP),max con-centration (MAX Conc),area under curve (AUC),MAX slope were calculated.The difference of these parameters be-tween malignant and benign breast lesions were calculated and compared by independent t test or nonparametric test. Additionally,ROC analysis was used to assess the diagnostic value of these quantitative and semi-quantitative parameters. Results:The values of Ktrans Max (t =2.228,P =0.033), Ktrans Mean (t =4.092,P <0.001),Kep Mean (Z =2.422, P =0.015),Kep Max (t =2.240,P =0.031),MAX Conc Max (t =3.256,P =0.002),MAX Conc Mean (t =3.460,P =0.001),AUC Max (t =2.250,P =0.034),AUC Mean (t =2.861,P =0.007),MAX Slope Max (t =2.478,P =0.018),MAX Slope Mean (t =2.226,P =0.032)were statistically significantly higher in malignant lesions than those in benign lesions,while the TTP Min (t =-5.145,P <0. 001),TTP Mean (t =-3.818,P <0.001)were obviously shortened in malignant lesions than those in benign lesions. The ROC analysis showed that the Ktrans Mean,Kep Mean,TTP Min,MAX Conc Max,AUC Mean,and MAX Slope Max were valuable in diagnosing malignant lesion from benign lesion.Adopting the maximum Youden’index as the cut-off val-ue,if the cut-off value of Ktrans Mean,Kep Mean,TTP Min,MAX Conc Max,AUC Mean,and MAX Slope Max were 0. 1137,0.0258,0.9065,0.1059,0.1640,and 0.1502 respectively,the sensitivity were 73.33%、73.33%、91.67%、76. 67%、80.00%、80.00%,and the specificity were 100%,83.30%,70.00%,91.70%,66.70%,75.00%,respective-ly.Conclusion:The differential diagnosis of benign and malignant breast lesions in human breast by quantitative and semi-quantitative parameters based on DCE-MRI is of clinical value and relatively high efficiency.It is a promising method as a noninvasive evaluation for microcirculation of breast tumor.