计算机模式识别法在甲状腺肿瘤患者血清微量元素谱研究中的应用  

Application of Computer Pattern Recognition Technique in Study on Trace Elements in Serum of Patients with Thyroid Tumors

作　　者：曾维钦 李增禧[2] 梁业成[2] 李小梁[2] 盛少禹[2] 余煜棉[3] 丁少群 陈超 

机构地区：[1]广州市天河区妇幼保健院,广州510630 [2]中国广州分析测试中心,广州510070 [3]广东工业大学化工系,广州510090

出　　处：《广东微量元素科学》1999年第2期19-24,共6页Trace Elements Science

摘　　要：用马氏距离判别法对53例甲状腺肿瘤患者和58例健康人血清中Ba、Cr、Mg、Mn、Mo、Sr、Zn、Fe、Cu和Ca等元素含量进行了分类研究,分类准确率达92.8％.研究结果表明,甲状腺肿瘤病变可能与人体中Mo、Mn、Fe、Mg、Ca和Sr含量异常有关.文中还计算了患者和健康人血清元素间的相关系数,并讨论了其差异性.The content of Ba, Cr, Mg, Mn, Mo, Sr, Zn, Fe, Cu and Ca in serum of 53 cases with thyroid tumors and 58 normal controls was studied by Mahalanobis distance distinguish method. The accuracy rate of classification was 92. 8%. The study showed that the occmce of thyroid tumors was related to the abnormal content of Mo, Mn, Fe, Mg, Ca and Sr in human boy. The related coefficients between trace elements in serums of patients and normal controls were calculated and their differences werediscussed.

关 键 词：计算机模式识别 甲状腺肿瘤 血清 微量元素 

分 类 号：R736.1[医药卫生—肿瘤] TP391.4[医药卫生—临床医学]