摘要
本研究采用多因素分析方法,定量研究了由核磁共振测量的正常组、病例组脑部形态结构,建立了诊断判别函数,评判符合率99.71％;进行了指标聚类,将性质相似的指标归为同类,独立性较强的指标自成一类;按年龄进行有序样品聚类,分析正常人脑部年龄特征,此项研究为探讨正常人和橄榄桥脑小脑萎缩病人的脑部形态特征提供了数量化依据。
In this paper, inultivariate analysis methods were used to quantitatively study the brain morphologic feature data measured by magnetic resonance imaging. First, the discriminant functions were estimated and discriminatory coincidence rate was 99. 7%. Second, cluster analysis method was used to classify simitar character indexes. Third, the ordered sample cluster analysis method was used to analyse age feature of normal brain and divide 5 age stage. This study provided quantitative information for research normal and disease brain morphologic feature.
出处
《数理医药学杂志》
1994年第3期237-240,共4页
Journal of Mathematical Medicine
关键词
聚类分析
NMR
脑
形态
结构
Discriminant function Cluster analysis Magnetic resonance imaging
