【摘要】 目的:探讨三维动脉自旋标记(3D-ASL)灌注成像对高、低级别胶质瘤鉴别诊断价值。方法:选择经手术及病理证实的脑胶质瘤35例,其中高级别胶质瘤22例,低级别胶质瘤13例,术前应用3.0 T MRI行常规MRI扫描、3D-ASL灌注成像及MRI增强检查,对感兴趣区脑血流量(CBF)进行测量和分析,数据分析使用SPSS 19.0软件。结果:高级别胶质瘤实性区域呈高灌注,低级别胶质瘤呈低、等灌注;高级别胶质瘤的平均血流量(mCBF)和相对血流量(rCBF)明显高于低级别胶质瘤(P<0.001)。根据ROC曲线分析,mCBF鉴别高、低级别胶质瘤的最佳临界值为57.8 mL/(min·100 g),敏感度和特异度分别为81.8%和100%。rCBF鉴别高、低级别胶质瘤的最佳临界值为89.7,敏感度和特异度分别为95.5%和84.6%。mCBF与rCBF的ROC曲线下面积差异无统计学意义(0.962 vs 0.913,P>0.05)。结论 :3D-ASL可以用于评估脑胶质瘤微血管灌注情况,mCBF及rCBF分别以57.8 mL/(min·100 g)及89.7作为最佳临界值,对胶质瘤的术前分级诊断具有较高的敏感度及特异度。更多还原

【Abstract】 Objective To study the diagnostic value of the three-dimensional Arterial Spin Labeling（3D-ASL）technique in high-gradeand low-grade gliomas. Methods Thirty-five cases of histologically proven gliomas were recruited in the study including 22 cases with high-grade gliomas and 13 with low-grade gliomas.Patientswereroutinelyexaminedby 3.0 T MRI including conventional MRI, 3D-ASL and enhanced MRI scan beforeoperation. In the regions of interest, maps of cerebral blood flow（CBF） were computed and analyzed. Statistical analysis was performedby using SPSS 19.0. Results In the patients, ASL technique was demonstrated to have higher perfusion values in imaging of high-grade gliomas, lower and equalvalues in imaging of low-grade tumors. Mean cerebral blood flow（mCBF） and relative cerebral blood flow（rCBF）values of high-grade gliomas were significantly higher than those of low-grade tumors. According to the ROC curve analysis, the best threshold mCBF distinguishing high-gradegliomasandlow-gradegliomas was 57.8 ml·min^-1·（100 mg）^-1, the sensitivity and the specificity rate were 81.8% and 100% respectively. The best critical rCBF value to differentiate these two kinds of tumors was 89.7, the sensitivity and the specificity rate were 95.5% and 84.6% respectively. Conclusions The 3D-ASL is a suitable method to evaluate microvascular perfusion. 57.8mL·min^-1·（100mg）^-1 of mCBF and 89.7 of rCBF are the best threshold to identify high and lowgrade gliomas effectively with higher diagnostic sensitivity and specificity.更多还原