摘要
目的探讨单指数、双指数、拉伸指数模型扩散加权成像(diffusion weightedimaging,DWI)对胶质瘤术前分级的诊断性能,评价灌注参数与血流量(cerebral bloodflow,CBF)的相关性。材料与方法收集经手术病理证实为脑胶质瘤的病例34例,包括低级别17例,高级别17例。受试者术前均行常规扫描、3D动脉自旋标记成像(threedemensional arterial spin labeling,3D ASL)及多b值DWI序列扫描,利用FUNCTIONTOOL后处理软件中的单指数模型计算得到表观扩散系数(apparent diffusion coefficient,ADC),利用双指数模型得到D、D*、f值,利用拉伸指数模型得到分布扩散系数(distribute diffusion coefficient,DDC)、水分子各向异性(alpha,α),利用3D ASL后处理软件得到CBF值。所得各参数利用SAS9.4统计软件进行分析。结果高级别胶质瘤ADC、D、f、DDC、ɑ值低于低级别胶质瘤,D*值高于低级别胶质瘤,P值均<0.05;全部胶质瘤D*与CBF呈正相关(r=0.51,P<0.05);受试者工作特性(receiver operatingcharactsteristic,ROC)曲线分析得出:ɑ、D、DDC、ADC值的曲线下面积(area undercurve,AUC)分别为0.969、0.965、0.961、0.957,其诊断阈值分别为0.88×10^-3mm^2/s、0.86×10^-3mm^2/s、1.10×10^-3mm^2/s、1.11×10^-3mm^2/s。结论单指数、双指数、拉伸指数模型DWI为胶质瘤术前提供更多诊断信息,ɑ、D、DDC、ADC值在鉴别高、低级别胶质瘤均有较高价值,且灌注参数D*与CBF呈正相关。
Objective: To explore the diagnostic performance of mono-exponential model,biexponential model and stretched-exponential model DWI for preoperative grading of gliomas, and evaluate the correlation between perfusion parameters and cerebral blood flow (CBF). Materials and Methods: Thirty-four patients with pathological confirmed glioma including 17 low-grade glioma patients and 17 high-grade glioma patients.Before operation,the subjects underwent coventionional MRI,three demensional arterial spin labeling (3D ASL) and multi-b value DWI;by using the mono-exponential model of FUNCTION TOOL software the apparent diffusion coefficient (ADC) was calculated;using bi-exponential model the fast ADC (D*), fraction of perfusion (f) and slow ADC (D) were calculated,using the stretched-exponential model the distribute diffusion coefficient (DDC) and alpha (α) were calculated,the CBF was obtained by 3D ASL software.The values obtained by FUNCTION TOOL software were analyzed by SAS9.4 statistical software.Results: The mean ADC,D,f,DDC,and ɑ values in high-grade glioma were lower than low-grade glioma,and the D* value was higher in highgrade glioma,P<0.05;there was positive correlation between D* and CBF of all the subjects (r=0.51,P<0.05);the receiver operating charactsteristic curve (ROC) analysis showed that the area under curve (AUC) of α,D,DDC and ADC values were 0.969,0.965,0.961,0.957 respectively,and the diagnostic thresholds of ɑ,D,DDC,and ADC were 0.88×10^-3 mm^2/s, 0.86×10^-3 mm^2/s,1.10×10^-3 mm^2/s,and 1.11×10^-3 mm^2/s,respectively.Conclusions: Monoexponential model,bi-exponential model and stretched-exponential model DWI can provide more diagnostic information of glioma before surgery,and ɑ,D,DDC,ADC values are of high value in identifying high- and low-grade gliomas.The perfusion parameter D* was positively correlated with CBF.
作者
郝凤玲
吴慧
牛广明
HAO Fengling;WU Hui;NIU Guangming(Department of Radiology,Affiliated Hospital of Inner Mongolia Medical University,Huhhot010000,China)
出处
《磁共振成像》
CAS
2019年第6期401-405,共5页
Chinese Journal of Magnetic Resonance Imaging
关键词
单指数模型
双指数模型
拉伸指数模型
神经胶质瘤
磁共振成像
mono-exponential model
bi-exponential model
stretched-exponential model
glioma
magnetic resonance imaging