全容积表观扩散系数直方图鉴别颅内孤立性纤维瘤/血管外皮细胞瘤与非典型脑膜瘤的价值

Value of Apparent Diffusion Coefficient Histogram in Differentiating Solitary Fibrous Tumor/Hemangiopericytoma and Anaplastic Meningioma

目的探讨全容积表观扩散系数(ADC)直方图参数在颅内孤立性纤维瘤/血管外皮瘤(SFT/HPC)与非典型脑膜瘤(AM)鉴别中的价值。方法回顾性分析经手术病理证实的20例SFT/HPC与19例AM的临床资料及术前MRI影像学资料。参照T1WI增强扫描图像,在ADC图上利用Firevoxel软件勾画感兴趣区(ROI),生成ADC直方图,提取多个ADC直方图参数,包括ADC最小值(ADC_(min))、ADC平均值(ADC_(mean))、ADC最大值(ADC_(max))、 ADC第10百分位数(ADC_(10%))、ADC第25百分位数(ADC_(25%))、ADC第50百分位数(ADC_(50%))、ADC第75百分位数(ADC_(75%))、ADC第90百分位数(ADC_(90%))、偏度系数(skewness)、峰度系数(kurtosis)及熵(entropy)。统计分析SFT/HPC与AM组中的各参数差异及其诊断效能。结果利用全容积ADC直方图提取的ADC_(mean),ADC_(max),ADC_(50%),ADC_(75%),ADC_(90%)及kurtosis在SFT/HPC与AM间均有统计学意义(P<0.05)。ADC_(mean)具有最高的鉴别诊断价值,SFT/HPC组的ADC_(mean)值为(1.176±0.386)×10^(-3)mm^(2)/s大于AM组(0.863±0.127)×10^(-3)mm^(2)/s, ADC_(mean)的受试者工作特征曲线(ROC)曲线下面积(AUC)为0.795。ADC_(mean)取1.048×10^(-3)mm^(2)/s作为截断点,诊断灵敏度为70.0%,特异度为89.5%。结论全容积ADC直方图分析法对鉴别SFT/HPC与AM具有一定价值,其中以ADC_(mean)的诊断价值最高。

Objective To investigate the value of Apparent Diffusion Coefficient(ADC) histogram in differentiating intracranial solitary fibroma/hemangiopericytoma(SFT/HPC) from atypical meningioma(AM). Methods The clinical data and preoperative MRI imaging data of 20 cases of SFT/HPC and 19 cases of AM confirmed by surgery and pathology were retrospectively analyzed. Refer to the T1 enhanced image, the Firevoxel software was used to outline the region of interest(ROI) on the ADC map, generate the ADC histogram, and record multiple ADC histogram parameters, including ADC_(mean),ADC_(min),ADC_(max),ADC_(10%),ADC_(25%),ADC_(50%),ADC_(75%),ADC_(90%),skewness, kurtosis and entropy. Statistical analysis of the differences in parameters between SFT/HPC and AM groups and their diagnostic efficacy. Results ADC_(mean),ADC_(max),ADC_(50%),ADC_(75%),ADC_(90%) and kurtosis extracted by full-volume ADC histogram were statistically significant between SFT/HPC and AM(P<0.05).ADC_(mean) has the highest differential diagnosis value. The ADC_(mean) value of the SFT/HPC group is(1.176±0.386)×10^(-3)mm^(2)/s greater than the AM group(0.863±0.127)×10^(-3)mm^(2)/s, which is below the ADC_(mean) ROC curve The area is 0.795.ADC_(mean) takes 1.048 as the cut-off point, the diagnostic sensitivity is 70.0%,and the specificity is 89.5%. Conclusion The full-volume apparent diffusion coefficient histogram analysis method has certain value in distinguishing SFT/HPC and AM,and ADC_(mean) has the highest diagnostic value.