摘要
目的探讨联合DWI纹理分析及常规MRI特征在鉴别颅内孤立性纤维瘤/血管外皮瘤(SFT/HPC)和不典型脑膜瘤(AM)中的应用价值。方法选取本院经手术及病理证实的14例SFT/HPC患者和29例AM患者常规MRI征象及DWI纹理特征。在ADC图像上勾画整个肿瘤的实质区作为感兴趣区容积(VOI),利用Omni-Kinetics软件进行纹理分析。采用χ^(2)检验、t检验、Logistic回归和受试者工作特征曲线(ROC)等比较常规MRI特征、DWI纹理特征以及联合应用对鉴别SFT/HPC和AM的诊断效能。结果SFT/HPC及AM两组肿瘤常规MRI特征及DWI纹理特征有显著性差异(P<0.05)。其中硬膜附着类型和DWI均匀性为鉴别的独立危险因素,ROC曲线下面积(AUC)分别为0.775和0.815,联合两者鉴别的AUC为0.938。结论DWI纹理分析及常规MRI特征可以鉴别SFT/HPC及AM,两者联合可以提高诊断效能。
Objective To assess the value of the texture analysis(TA)of DWI and conventional MRI features,alone and in combination,in differentiating intracranial solitary fibrous tumor/hemangiopericytoma(SFT/HPC)from atipical meningioma(AM).Methods Pathologically,confirmed intracranial SFT/HPC(n=14)and AM(n=29)were retrospectively collected and their conventional MRI features were analyzed.The VOI(volume of interest)covering the entire tumor was drawn on the original ADC maps and the corresponding texture parameters were calculated.Theχ^(2) test,unpaired student's t-test,Logistic regression and receiver operating characteristic(ROC)curve analysis were used to analyze the diagnostic performance of each parameter and their combination for differentiating intracranial SFT/HPC from AM.Results The conventional MRI features and the parameters of TA derived from DWI were significantly different between SFT/HPC and AM(P<0.05),of which the type of dural attachment and the uniformity of DWI had the highest area under the ROC curve(AUC)(0.775 and 0.815).The AUC of the combination of DWI uniformity and type of dural attachment in differentiation of SFT/HPC and AM was 0.938.Conclusion TA of DWI and conventional MRI features are helpful to differentiate SFT/HPC from AM.Combining the TA of DWI and conventional MRI features within a statistical classifier can help improve the differentiation of SFT/HPC and AM.
作者
陈小田
苏春秋
潘婷
张璇
鲁珊珊
吴飞云
施海彬
洪汛宁
CHEN Xiaotian;SU Chunqiu;PAN Ting;ZHANG Xuan;LU Shanshan;WU Feiyun;SHI Haibin;HONG Xunning(Department of Radiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, China)
出处
《医学影像学杂志》
2022年第5期742-747,共6页
Journal of Medical Imaging
关键词
颅内孤立性纤维瘤/血管外皮瘤
不典型脑膜瘤
纹理分析
磁共振成像
Intracranial solitary fibrous tumor/hemangiopericytoma
Atipical meningioma
Texture analysis
Magnetic resonance imaging
