基于近红外漫反射光谱和多元数据分析的黄芩质量标准的快速评价方法研究

目的: 采用近红外漫反射光谱法结合多元数据分析对黄芩中水分、灰分、醇浸出物和黄芩苷进行快速的同步定量分析,同时建立黄芩质量标准的快速评价方法. 方法: 采集85批不同来源黄芩样品的近红外光谱数据,按照中国药典2010年版一部黄芩项下要求测定黄芩水分、灰分、醇溶性浸出物和黄芩苷含量,比较研究不同预处理方法对模型预测性能的影响,并将窗口移动最小二乘法(MWPLS)用于优化建模光谱区间. 结果: 黄芩苷、水分、灰分和醇溶性浸出物4个验证模型的相关系数(r)分别为0.945 1、0.895 5,0.930 8和0.915 0,相应的预测误差均方根(RMSEP)分别为1.29、0.669、0.492和3.47. 结论: 该方法可为建立黄芩质量标准的快速评价方法打下良好的基础.

Objective: To apply near infrared(NIR)diffuse reflectance spectroscopy and multivariate analysis to develop a simultaneous quantification method for water,ash,ethanol-soluble extract and baicalin,and establish a fast assessment on quality standard of Radix Scutellariae. Methods: NIR data of 85 samples from different origins were collected.The contents of water,ash,ethanol-soluble extract and baicalin were determined according to the Pharmacopoeia of the People's Republic of China(2010).The performance of the pretreatment methods and spectral region selection were compared and evaluated and the method of moving window partial least square(MWPLS)was used to optimize modeling spectral region. Results: The result showed that the correlation coefficients of the validation models were 0.945 1 for baicalin,0.895 5 for water,0.930 8 for ash and 0.915 0 for ethanol-soluble extract,respectively.In addition,the root mean square errors of prediction(RMSEPs)were 1.29 for baicalin,0.669 for water,0.492 for ash and 3.47 for ethanol-soluble extract,respectively. Conclusion: The proposed method is fast and non-destructive,and the result can provide a good foundation for the establishment of the rapid evaluation method of quality standard of Radix Scutellariae.