Fisher成分分析法用于瓜蒌及其炮制品的化学指纹特征相似性的研究

目的：应用Fisher成分分析法（FCA）对瓜蒌及其炮制品的类别进行分析。方法采用HPLC-PDAD建立瓜蒌及其炮制品的指纹图谱，以槲皮素为内参峰进行数字化处理，得到瓜蒌及其炮制品的标准指纹图谱。以FCA提取瓜蒌及其炮制品标准指纹图谱中的化学信息，进行质量模式分类，并与标准指纹图谱及主成分分类模式进行对比。结果 FCA能准确地将瓜蒌及其炮制品分为2类。结论采用FCA成分分析法可提取指纹图谱中隐含的化学信息，进行准确的类别分析。

Application of Fisher component analysis to study chemical fingerprint feature similarities of Fructus Trichosanthis and its processed products

Objective To explore whether the basic principle of Fisher component analysis(FCA)can be used in the catego?ry analysis of Fructus Trichosanthis fruit and its processed products. Methods Their fingerprints were established by using HPLC-PDAD,and the standard fingerprints of them were obtained by digitizing with quercetin as the internal reference peak. Then,their chemical fingerprint information was extracted by FCA and classified by quality model,and the classification results were compared with those classified by principal component analysis and standard atlas analysis. Results Fructus Trichosanthis and its processed products can be accurately divided into two categories by FCA. Conclusion FCA can extract the implied chemical information in fin?gerprints,and analyze them accurately.