基于直方图分析法定量评估DCE-MRI在脑胶质瘤分级诊断中的价值

Histogram analysis for quantitative dynamic contrast enhanced MRI in grading of glioma

目的通过直方图法分析磁共振动态对比增强(dynamic contrast enhanced magnetic resonance imaging,DCE-MRI)定量评估脑胶质瘤分级的价值。材料与方法 120例脑胶质瘤患者中,低级别(Ⅰ、Ⅱ级)28例,高级别(Ⅲ、Ⅳ级)92例。采用改良的extended Toft双室模型,获得DCE渗透性参数K^(trans)、K_(ep)、V_e、V_p和灌注参数时间浓度曲线对比剂首过的曲线下面积(initial area under gadolinium concentration time curve,IAUGC)、脑血流量(cerebral blood flow,CBF)、平均通过时间(mean transit time,MTT)。应用非配对t检验和Mann-Whitney U检验比较不同级别胶质瘤DCE直方图参数的差异,应用受试者工作特征(receiver operating characteristic,ROC)曲线对比分析各不同参数的诊断效能、敏感度和特异度及相应的诊断阈值。结果高、低级别胶质瘤K^(trans)、K_(ep)、V_e和IAUGC、MTT的差异有统计学意义(P<0.05)。ROC曲线分析显示,IAUGC和V_e的诊断效果最佳(P<0.05)。与平均值相比,K^(trans)的第75百分位数诊断效能升高最明显,IAUGC的第95百分位数的诊断效能最高。结论基于直方图分析法的DCE参数对定量评估脑胶质瘤的分级具有重要的临床参考价值。

Objective:This study aimed to quantitatively evaluate the value of dynamic contrast enhanced magnetic resonance imaging(DCE-MRI)in grading of glioma using histogram analysis.Materials and Methods:One hundred and twenty glioma patients were included with 28 low grade gliomas(LGG,gradeⅠand gradeⅡ)and 92 high grade gliomas(HGG,gradeⅢand gradeⅣ).Permeability parameters,including Ktrans,Kep,Ve,Vp,and perfusion parameters,including IAUGC,CBF and MTT,were derived from DCE-MRI by applying the modified two-compartment extended toft model.Unpaired t test and Mann-Whitney U test were used to compare the difference between DCE-MRI histogram parameters of different grade gliomas.The diagnostic efficiency,sensitivity,specificity and the corresponding diagnostic threshold value of different parameters were then compared by ROC.Results:There were significant differences of Ktrans,Kep,Ve,IAUGC and MTT between LGG and HGG(P<0.05).The results of ROC showed that IAUGC and Ve had the best diagnostic performance(P<0.05).The diagnostic efficiency of the 75th percentile of Ktrans had the most significant increase compare with the mean value.The 95th percentile of IAUGC had the best performance.Conclusions:Histogram analysis for DCE-MRI provides valuable information for glioma grading.