基于磁共振图像的脑肿瘤自动识别与分析  

Automatic Identification and Analysis of Cerebral Tumors Based on MR Images

作　　者：高宏建[1] 薛艳青[1] 吴水才[1] 艾海明[1] 杨春兰[1] 白燕萍[1] 盛磊[1] 刘伟[2] 

机构地区：[1]北京工业大学生命科学与生物工程学院,北京100124 [2]北京大学人民医院放射科,北京100044

出　　处：《北京工业大学学报》2012年第6期955-960,共6页Journal of Beijing University of Technology

基　　金：北京市自然科学基金资助项目(11002063);北京工业大学青年基金项目(X10159992010022)

摘　　要：介绍了基于磁共振图像的脑肿瘤自动识别及三维分析方法.通过支持向量机方法对脑部磁共振图像建立数学模型,实现脑肿瘤的自动识别.利用图像重建获得脑肿瘤的三维模型,并通过形态学分析提取相关特征参数.实验结果表明,该方法可从脑磁共振图像中有效识别并重建脑肿瘤,从而为临床进行射频治疗手术提供可靠依据和客观评价.Automatic identification and analysis methods of cerebral tumors are introduced. The mathematical model of brain MRI is built by support vector machines, thereby achieving automated identification of cerebral tumors. Then, 3D surface models of cerebral tumors can be obtained through image reconstruction and associated 3D physical parameters can be extracted based on the morphologic analysis. Experimental results show that the method can effectively identify 3D cerebral tumors and thus provides reliable clinical evidence and objective evaluation for the radiofrequency thermocoagulation operation of cerebral tumors.

关 键 词：脑肿瘤 自动识别 射频消融 

分 类 号：R318[医药卫生—生物医学工程]