一类用化学指纹特征鉴别中药材真伪的方法

Identifying the Authenticity of Chinese Medicinal Material with Chemical Fingerprint Features

提出用 Fisher因子表征中药材质量模式 ,根据类间距离最大、类内距离最小原则 ,计算色谱图中各峰的权值 ,并以各峰权值为变换向量从各峰面积值中计算出 Fisher因子 ,用于中药材质量模式分类 .以红参主根的真伪鉴别为实例 ,从定量评价、目视比较和聚类分析等方面将 Fisher因子、主成分及指纹峰进行分类效果对比研究 .结果表明 ,Fisher因子能更好地表征红参化学模式特征 ,分类准确性最高 .因此 ,用本文方法提取的 Fisher因子是一类能表征中药材质量模式的化学指纹特征参量 ,适宜用作鉴别中药材真伪 .

A method for identifying the authenticity of Chinese Medicinal Material(CMM), using Fisher factor(FF) as chemical fingerprint features, is proposed. According to the principle of maximizing the distance among classes and minimizing the distance within classes, the weights of different peaks in chromatograms are calculated, and then a transformation vector is composed by these weights. FF can be obtained by multiplying the transformation vector with the vector composed of the area values of every peak selected based on users threshold. As a typical example, identifying the authenticity of the main root of red ginseng was used to verify the effectivity of the proposed method. Comparing FFs with principal components and chromatographic peaks from three aspects of the quantitative evaluation, visual analysis and cluster analysis, it is shown that FF has the highest accuracy in identifying the authenticity of the main root of red ginseng. Therefore, as a chemical fingerprint feature, FF can be used to identify the authenticity of CMM.