摘要
目的:研究ADC灰度直方图鉴别侧脑室内中枢神经细胞瘤和室管膜瘤的价值。方法:回顾性分析本院经手术病理证实的发生于侧脑室内的10例室管膜瘤和14例中枢神经细胞瘤,选取两组肿瘤ADC横轴面图像最大层面,并用MaZda软件勾画兴趣区(ROI)并进行直方图分析,对比两组ADC灰度直方图参数,包括平均值(Mean)、方差(Variance)、偏度(Skewness)、峰值(Kurtosis)、第1百分位数(Perc.01%)、第10百分位数(Perc.10%)、第50百分位数(Perc.50%)、第90百分位数(Perc.90%)、第99百分位数(Perc.99%),绘制ROC曲线分析以上参数对室管膜瘤和中枢神经细胞瘤的鉴别诊断效能。结果:通过ADC灰度直方图分析得到的9个纹理参数中,平均值、方差、Perc.01%、Perc.10%、Perc.50%、Perc.90%、Perc.99%差异有统计学意义(P<0.05)。其中,Perc.99%的曲线下面积最大(0.979),其敏感度和特异度分别为100%、92.9%,具有较高的鉴别效能。结论:ADC灰度直方图分析可用于鉴别室管膜瘤和中枢神经细胞瘤。
Objective:To study the value of ADC histogram analysis for differential diagnosis between central neurocytoma and ependymoma in the lateral ventricle.Methods:Ten patients with ependymoma and 14 patients with central neurocytoma pathologically confirmed were analyzed retrospectively.Regions of interest(ROIs)in the ADC maps of two groups were drawn on maximum level of tumor and histogram analysis were undergone by using MaZda software.The parameters from ADC histogram included mean,variance,skewness,kurtosis,perc.01%,perc.10%,perc.50%,perc.90%and perc.99%.ROC curve analysis was used to compare the differential diagnosis efficiency of each parameter.Results:Through histogram analysis of nine parameters,seven parameters showed statistically significant differences(P<0.05),including mean,variance,perc.01%,perc.10%,perc.50%,perc.90%and perc.99%.The other parameters had no significant differences between the two groups(P>0.05).Perc.99%had the maximum area under the ROC curve of 0.979,with sensitivity of 100%,and specificity of 92.9%.It had high identification efficiency.Conclusion:The ADC histogram analysis can be a new method for the differential diagnosis between central neurocytoma and ependymoma.
作者
周健
程敬亮
陈晨
张勇
谢珊珊
ZHOU Jian;CHENG Jing-liang;CHEN Chen(Department of Magnetic Resonance,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450052,China)
出处
《放射学实践》
北大核心
2020年第3期325-328,共4页
Radiologic Practice
关键词
脑肿瘤
中枢神经细胞瘤
室管膜瘤
磁共振成像
直方图分析
Brain neoplasms
Central neurocytoma
Ependymoma
Magnetic resonance imaging
Histogram analysis
