摘要
本文尝试利用化学计量学技术从整体上探讨无机元素含量水平与中药四性的关系,对115味中草药的32种元素的三电极直流等离子体原子发射光谱数据应用模式识别方法进行了系统的多变量统计分析。以中药四性理论为依据的三类药物(温热药、平性药、寒凉药)中,有监督模式识别方法之一——线性判别法(LDA)的正确判断能力高达70.43%,显示了这115味中草药的数值分类结果和以中药四性理论为依据自然分类的良好的一致性;2个无机元素含量均值中,寒凉药类有84.37%高于整体,而温热药类和平性药类分别有90.62%和84.37%低于整体,进一步证明了中药无机元素含量水平是决定中药四性的物质基础之一,揭示了微量元素含量和中药疗效之间存在的相关性。
The relationship between four characteristics and the amount of trace elements contained in Chinese herbs was explored as a whole. 32 kinds of trace elements in 115 Chinese herbs were determined by three-electrode inductive coupled plasma torch emission atomic spectrometer and the data sets were analysed by a commercial package of programs, PARVUS. The Linear Discriminant Analysis, one of Supervised Pattern Recognition approaches, was used and showed the excellent classification ability for the data sets categorized according to their traditional four characteristics. For the mean values of 32 trace elements, the 84.37% of 115 herbs with cool characteristics is higher than the whole, and 90.62% and 84.37% for the neutral and warm. The correct classifications in three categories are 70.34%. It was certified that the amount of trace elements is the substantial basics of the four characteristics in Chinese medicinal herbs, and the correlation between the amount of the trace elements and the curative effects in Chinese medicinal herbs exists.
出处
《中国药科大学学报》
CAS
CSCD
北大核心
1992年第6期348-353,共6页
Journal of China Pharmaceutical University
关键词
微量元素
中草药
四性
化学计量学
Trace elements
Four Characteristics of herbs
Chemometrics
Supervised Pattern Recognition
