基于激光诱导击穿光谱技术的三七产地及道地研究  

Study on Geoherbalism and Origin Identification of Sanqi(Panax Notoginseng)Based on Laser-induced Breakdown Spectroscopy

作　　者：车晓青 夏子怡 李坤玉 李文[4] 刘晓娜[1] CHE Xiaoqing;XIA Ziyi;LI Kunyu;LI Wen;LIU Xiaona(College of Traditional Chinese Medicine,Binzhou Medical University,Yantai 264003,Shandong,China;Shandong Xianglong Pharmaceutical Research Institute Co.,Ltd.,Yantai 264003,Shandong,China;Hebei Baicaokangshen Pharmaceutical Co.,Ltd,Shenzhou 053800,Hebei,China;Lanzhou University,Lanzhou 730000,Gansuu,China)

机构地区：[1]滨州医学院中医学院,山东烟台264003 [2]山东祥隆医药研究院有限公司,山东烟台264003 [3]河北百草康神药业有限公司,河北深州053800 [4]兰州大学,甘肃兰州730000

出　　处：《中华中医药学刊》2024年第8期46-49,I0014,共5页Chinese Archives of Traditional Chinese Medicine

基　　金：国家自然科学基金项目(81903796);国家重点研发计划项目(2019YFC1711204);山东省自然科学基金面上项目(ZR2022MH159);中国中医科学院科技创新工程项目(C12021A03704)。

摘　　要：目的从元素角度出发,采用激光诱导击穿光谱(laser-induced breakdown spectroscopy,LIBS)结合化学计量学方法研究三七的道地性及产地鉴别。方法采集云南文山三七和曲靖三七的LIBS光谱数据,进行数据归一化处理,利用偏最小二乘判别分析(PLS-DA)方法判别三七产地,并采用VIP值筛选出对产地判别模型的贡献大的元素。结果基于LIBS技术和PLS-DA方法可以有效地进行云南文山三七和曲靖三七的产地判别。以450~840 nm全光谱为输入变量,PLS-DA判别模型训练集的敏感性、特异性和总判正率均为100%,预测集敏感性、特异性和总判正率分别为90.9%、90%和95%;以VIP值筛选的特征变量作为输入变量,PLS-DA判别分析模型的性能良好,训练集和预测集的敏感性、特异性和总判正率均为100%。表明VIP值筛选的特征变量可代表三七的产地信息。贡献度较大的元素分别为:Si、Mg、Ca、Na、K等。结论LIBS技术结合多变量分析方法能够快速实现云南文山三七和曲靖三七的道地性研究。所建立的方法具有分析快速、近似无损、样品前处理简单等特点,为中药产地和道地性研究提供新思路。Objective The genuineness and origin identification of Sanqi(Panax Notoginseng)were studied based on laser-induced breakdown spectroscopy(LIBS)combined with chemometrics.Method Partial least squares discriminant analysis(PLS-DA)was implemented to discrimate origin of Sanqi(Panax Notoginseng)from Wenshan and Qujing in Yunnan province.Then VIP value was used to screen the important elements contributing to the classification.Results Combined with PLS-DA,LIBS can discriminant Sanqi(Panax Notoginseng)of Wenshan and Qujing effectively.The PLS-DA model using complete spectra as input variables showed good classification performance.The specificity,sensitivity and total accuracy were all 100%in training set of PLS-DA model,and were 90.9%,90%and 95%in the validation set,respectively.The VIP value was used to screen the variables contributing to the classification of PLS-DA model.Furthermore,the other PLS-DA model was constructed using the screen the variables by VIP value as input variables,which had better discrimination performance.The down-selection of the dataset specifically focused on the major elements(Si,Mg,Ca,Na and K),devoting to the geoherbalism of Sanqi(Panax Notoginseng).Conclusion This work proposed a novel method for origin identification and geoherbalism study of Chinese medicine.

关 键 词：道地性 产地 三七 激光诱导击穿光谱 元素 

分 类 号：R282.71[医药卫生—中药学]