扩散峰度联合动态增强磁共振成像对乳腺癌不同中医证型的诊断价值

Prediction of the expression of TCM syndrome in breast cancer based on diffusion kurtosis imaging and dynamic contrast enhancement imaging

目的探讨扩散峰度成像(DKI)联合动态增强磁共振成像(DCE-MRI)对乳腺癌不同中医证型的诊断价值。方法选取本院经病理证实的乳腺癌患者170例,其中肝郁痰凝证88例,冲任失调证45例,正虚毒炽证37例,所有病例均行DKI及DCE-MRI检查。分析两组间的临床病例资料、常规MRI征象和DKI模型的平均扩散率(MD)、平均扩散峰度值(MK)及DCE-MRI模型的血管外细胞外间隙容积比(Ve)、速率常数(Kep)和容量转移常数(Ktrans)值;并运用受试者工作特征曲线(ROC)分析诸定量参数的诊断效能。结果肝郁痰凝证、冲任失调证及正虚毒炽证病灶形态及强化方式差异有统计学意义(P=0.003;P=0.027)。肝郁痰凝证组的MD值小于冲任失调证组及正虚毒炽证组,差异有统计学意义(P<0.01);肝郁痰凝证、冲任失调证及正虚毒炽证组的MK值逐渐减小(P<0.001),且两两之间比较差异有统计学意义(P<0.05);正虚毒炽证组的Ktrans及Kep值最大,其次是冲任失调证组,肝郁痰凝证组最小(P<0.001),两两之间比较差异有统计学意义(P<0.05)。常规MRI模型中病变形态特征及强化方式对鉴别正虚毒炽证与非正虚毒炽证乳腺癌有意义,ROC曲线下面积为0.772。DCE-MRI模型的Ktrans值对正虚毒炽证的诊断效能最高,AUC均大于其他定量参数(Z=2.729~3.852,P<0.05)。DKI模型(MD+MK)的AUC为0.745;DCE-MRI模型(Ktrans+Kep)的AUC为0.902,大于DKI及常规MRI模型,差异具有统计学意义(Z=3.002,P=0.003;Z=2.113,P=0.035)。DCE-MRI+DKI模型AUC为0.917,诊断效能高于DKI及常规MRI模型(Z=3.834,P<0.001;Z=2.556,P=0.011);DCE-MRI+DKI模型的准确度91.8%,均高于DCE-MRI、DKI及常规MRI单一模型。结论DKI和DCE-MRI模型可用于中医证型乳腺癌的鉴别诊断,其中DCE-MRI模型的Ktrans值诊断效能最高,DKI联合DCE-MRI模型对乳腺癌中医证型的鉴别能力优于单一模型。

Objective To investigate predictive value of diffusion kurtosis imaging(DKI)combined with dynamic contrastenhanced magnetic resonance imaging(DCE-MRI)for the expression of TCM syndrome in breast cancer.Methods A total of 170 patients with breast cancer confirmed by pathology were collected,including a total of 88 cases of liver depression phlegm coagulation syndrome,45 cases of Chongren dysfunction syndrome,and 37 cases of right vacuity exuberant toxin syndrome.DKI and DCE-MRI scanning were performed in all patients.The clinicopathological characteristics and conventional MRI features and mean diffusion rate(MD),mean kurtosis value(MK),extravascular extracellular space distribute volume per unit tissue volume(Ve),rate constan(t Kep)and volume transfer constan(t Ktrans)were analyzed.The ROC was plotted to compare the diagnostic efficiency.Results There were statistical differences in the lesion shape and enhancement mode among the liver depression and phlegm coagulation syndrome,Chongren dysfunction syndrome,and right vacuity exuberant toxin syndrome group(P=0.003;P=0.027).The MK values of the liver depression phlegm coagulation syndrome,Chongren dysfunction syndrome,and right vacuity exuberant toxin syndrome groups gradually decreased(P<0.001),and the difference between the two groups was statistically significant(P<0.05).The Ktrans and Kep values were the highest in the right vacuity exuberant toxin syndrome group,followed by the Chongren dysfunction syndrome group,and the lowest in the liver depression phlegm coagulation syndrome group(P<0.001),with statistically significant differences between the two groups(P<0.05).The MD value of the liver depression phlegm coagulation syndrome group was lower than that of other TCM syndrome groups(P<0.01).The shape and enhancement mode of lesions in the conventional MRI model were significant to differentiate breast cancer with the syndrome of positive deficiency toxic incandescence from the syndrome of non-positive deficiency toxic incandescence.The area under the ROC curve was 0.772.The DKI model was MD+MK,with an AUC of 0.745;The DCE-MRI model was Ktrans+Kep,with an AUC of 0.902,which was greater than the DKI model and the conventional MRI model,and the difference was statistically significant(Z=3.002,P=0.003;Z=2.113,P=0.035).The AUC of the combined model of IVIM and DKI was 0.917,which was higher than that of the single model of DKI and conventional MRI(Z=3.834,P<0.001:Z=2.556,P=0.011).The accuracy of IVIM+DKI model was 91.8%,which were higher than those of DCE-MRI,DKI and conventional MRI single models,correspondingly.Conclusion The DKI and DCE-MRI models could be used to predict the expression of TCM syndrome in breast cancer.Ktrans had the highest diagnostic efficacy.DKI combined with DCE-MRI model was superior to single model in the prediction of the expression of TCM syndrome in the breast cancer.