摘要
分析了中药标准指纹图谱的传统构建方法,针对传统方法建立的中药标准指纹图谱在中药质量鉴别及评价上的局限性,进行了标准指纹图谱建立方法研究,提出了基于多参数距离融合聚类原理建立中药标准指纹图谱的方法。该方法通过计算中药多维多息图谱数据的各维特征聚类规则并融合在一起获得中药标准指纹图谱,实现了多种图谱信息的融合。采用黄芩数据进行了实验,实验结果显示依据聚类原理构建的中药标准指纹图谱更利于中药的分类。该方法是一种可行的中药标准指纹图谱构建方法。
The traditional methods for building standard TCM（ traditional Chinese medicine） fingerprints were analyzed, and aiming at the limitations of the TCM fingerprints constructed by traditional methods in TCM＇ s quality identifica- tion and evaluation, a novel method for construction of standard TCM fingerprints based on multi-parameter distance clusting theories was advanced. The new method builds standard TCM fingerprints by computing each feature cluste- ring rule of multi-dimensional print data of TCM and integrating them, to realize the fusion of multi-print informa- tion. The total culstering rule is the standard TCM fingerprint. The experiments were conducted by using Scutellaria data. The experimental results show that the standard TCM fingerprints constructed based on cluster theories are bet- ter than the traditional standard TCM fingerprints in TCM classification. It is a feasible method for construction of standard TCM fingerprints.
出处
《高技术通讯》
CAS
CSCD
北大核心
2014年第1期99-103,共5页
Chinese High Technology Letters
基金
中国博士后科学基金(2012M510722)
中国中医科学研究院项目(ZZ12001)
燕山大学博士基金(B692)
秦皇岛市科学技术研究与发展计划(201001A119)资助项目
关键词
中药指纹图谱
聚类分析
信息融合
可视化
TCM fingerprint, cluster analysis, information fusion, visualization
作者简介
女，1977年生，博士；研究方向：模式识别，中药指纹图谱；E—mail：ydcuijianxin@yahoo．com．cn
通讯作者。E—mail：hhghb@ysu．edu．cn
