计算机模式识别法对关节变形强直性脊柱炎微量元素谱的研究  

Study of Trace Element Table of Arthrosis Rigidity Rachitis by Computer Pattern Recognition

作　　者：刘俊[1] 余煜棉[2] 李燕玉[3] 

机构地区：[1]华中科技大学,武汉430074 [2]广东工业大学,广州510090 [3]中山医科大学,广州510090

出　　处：《广东微量元素科学》2001年第9期46-51,共6页Trace Elements Science

摘　　要：测定了关节变形强直性脊柱炎患者及健康人头发样本中的十五种微量元素含量 ,运用计算机模式识别技术研究两类样本的微量元素谱差异 ,选择Ca、Pb、Zn、Ti、Ni、Mg和Sr元素为特征参量 ,用马氏距离法对两类样本进行回判和预报 ,其准确率均高于 92 % ,高维空间分类成功 。The content of 15 kinds of trace elements in hair swatches which was collected from healthy people and sufferers with arthrosis rigidity rachitis were measeured.Computer pattern recognition technology was used to study the difference of the trace element table from the two kinds of swatches.Ca,Pb,Zn,Ti,Ni,Mg and Sr were selected as the character parameter.Mahalamobis distance decision method was used to judge and predict the swatches and the correct rate was above 92%.The research show that the evidente differences exist in the trace element table of the two kinds of people.

关 键 词：计算机模式识别 关节变形强直性脊柱炎 微量元素 头发 微量元素谱 

分 类 号：R593.23[医药卫生—内科学]