摘要
目的建立一种用近红外光谱(NIRS)鉴别川贝母及其伪品的方法。方法使用近红外光谱仪扫描贝母样品,对所得光谱进行预处理后,采用K值聚类法并结合偏最小二乘法对该图谱进行聚类分析。结果浙贝母、湖北贝母、平贝母、伊贝母的类型值分别为1.20、1.91、3.45、4.02,川贝母的类型值区间为5.06~5.96。结论所建NIRS法可有效区分川贝母和其他贝母,同时,也能初步将川贝母中的暗紫贝母、卷叶贝母、瓦布贝母、太白贝母、梭砂贝母和甘肃贝母等6个品种区分开来,可作为鉴别川贝母鳞茎的一种新方法。
OBJECTIVE To identify the bulbs of Fritillariae species by Near-infrared spectroscopy(NIRS).METHODS The full near-infrared spectra of Fritillariae bulbs samples was obtained using NIRS and analyzed by partial least squares(PLS)regression and K-means clustering analysis after the effects of spectral pre-processing and wavelength selection being investigated.RESULTS Five officinal Fritillariae species were successfully differentiated and characterized with the combination of NIRS,PLS regression and K-means clustering analysis.CONCLUSION The developed NIRS method is a promising method which is helpful to identify Fritillariae bulbs,but further studies still need to be carried out to verify the effectiveness of the developed model.
作者
周婷
付绍兵
谢慧敏
谢慧淦
叶利明
王曙
叶本贵
ZHOU Ting;FU Shaobing;TSE Wai Ming;TSE Kathy Gaun;YE Liming;WANG Shu;YE Bengui(West China School of Pharmacy,Sichuan University,Chengdu,Sichuan,610041 P.R.China;Nin Jiom Medicine Manufactory(H.K.)Limited,Hong Kong,China)
出处
《华西药学杂志》
CAS
CSCD
2021年第2期193-197,共5页
West China Journal of Pharmaceutical Sciences
基金
四川省科技计划项目资助(2017YFC1701801)
全国第四次中药资源普查项目资助(2018PC015)
四川大学校地合作项目(2019CDLZ-25)。
关键词
贝母
川贝母
近红外光谱法
化学计量学
偏最小二乘法回归
K值聚类法
聚类分析
真伪鉴别
Fritillariae bulbs
Fritillaria cirrhosa
Near infrared spectroscopy
Chemometrics
Partial least squares regression
K-means clustering method
Clustering analysis
Identification
