合成MRI联合常规MRI鉴别三阴性与非三阴性乳腺癌的价值  

Value of synthetic MRI and conventional MRI in identifying triple negative and non-triple negative breast cancer

作　　者：王兰梅 周智鹏[1] 徐列印[1] 林斌 余雪燕 王玲廷 WANG Lanmei;ZHOU Zhipeng;XU Lieyin;LIN Bin;YU Xueyan;WANG Lingting(Department of Radiology,Affiliated Hospital of Guilin Medical College,Guilin 541001,China)

机构地区：[1]桂林医学院附属医院放射科,桂林541001

出　　处：《磁共振成像》2024年第7期112-117,共6页Chinese Journal of Magnetic Resonance Imaging

基　　金：广西自然科学基金项目(编号:2022JJA141266);2022年桂林市技术应用与推广计划项目(编号:20220139-2)。

摘　　要：目的 探讨合成MRI(synthetic magnetic resonance imaging, syMRI)联合常规MRI鉴别三阴性乳腺癌(triple negative breast cancer, TNBC)与非TNBC的价值。材料与方法 对经病理证实的167例乳腺癌病变进行回顾性分析,TNBC 30例,非TNBC 137例,所有患者均接受syMRI检查,测量评估以下参数:纤维腺体组织、实质背景强化、病灶形态、边缘、直径、T2WI信号、表观扩散系数(apparent diffusion coefficient, ADC)、强化方式、时间-信号强度曲线、syMRI参数(T1pre、T2pre、PDpre、T1Gd、T2Gd、PDGd)。单因素、多因素分析比较TNBC组与非TNBC组参数,并建立三种预测模型,syMRI模型、常规MRI模型、联合模型(syMRI+常规MRI),采用受试者操作特征曲线及曲线下面积(area under the curve, AUC)分析各预测模型鉴别TNBC与非TNBC的效能,DeLong检验比较AUC差异。结果 TNBC组与非TNBC组间的T2WI信号、ADC值、边缘强化、T1pre、T2pre、T1Gd、T2Gd、ΔT2、ΔPD、rT2、ΔT2%、ΔPD%差异具有统计学意义(P<0.05)。多因素logistic回归分析显示,ADC、边缘强化、T2pre、T1Gd、T2Gd是TNBC诊断的独立预测因子(P<0.05),联合模型的诊断性能最高,AUC为0.932。结论 基于syMRI、常规MRI建立的预测模型有助于鉴别TNBC和非TNBC。Objective:To explore the value of synthetic MRI(syMRI)and conventional MRI in identifying triple negative breast cancer(TNBC)and nonTNBC.Materials and Methods:This retrospective study included 167 patients with pathologically proven breast cancer.There were 30 cases in the TNBC group and 137 cases in the nonTNBC group.All patients underwent syMRI examinations.The following parameters were measured and evaluated:fibroglandular tissue,background parenchymal enhancement,lesion shape,edge,diameter,T2WI signal,apparent diffusion coefficient(ADC),patterns of enhancement,time-signal intensity curve,syMRI parameters(T1pre,T2pre,PDpre,T1Gd,T2Gd,PDGd).Univariate and multivariate analysis were used to compare the parameters of TNBC group and nonTNBC group and three predictive models were established:syMRI model,conventional MRI model and Joint model(syMRI+conventional MRI).Receiver operator characteristic curve and area under the curve(AUC)were used to analyze the efficiency of each predictive model in distinguishing TNBC and non-TNBC.Then the DeLong test was used to compare the differences in AUC.Results:There were statistically significant differences(P<0.05)in T2WI,ADC,rim enhancement,T1pre,T2pre,T1Gd,T2Gd,ΔT2,ΔPD,rT2,ΔT2%andΔPD%between TNBC group and non TNBC group.Multivariate analyses showed that ADC,the presence of rim enhancement,T2pre,T1Gd and T2Gd were independent predictors for diagnosis of TNBC(P<0.05).Joint model had the highest diagnostic performance with an AUC of 0.932.Conclusions:The prediction model established based on syMRI and conventional MRI have value in identifying TNBC and nonTNBC.

关 键 词：三阴性乳腺癌 非三阴性乳腺癌 磁共振成像 合成磁共振成像 常规磁共振成像 

分 类 号：R445.2[医药卫生—影像医学与核医学] R737.9[医药卫生—诊断学]