摘要
目的联合采用近红外(NIR)和中红外(MIR)光谱技术,通过数据融合方式建立姜半夏及其伪品姜虎掌南星的快速鉴别方法。方法收集22批姜半夏和14批姜虎掌南星样品,采集其NIR和MIR光谱数据,以偏最小二乘判别分析(PLS-DA)算法分别建立单一光谱数据和融合数据的姜半夏及其伪品的判别模型,以分类准确率对不同PLS-DA模型的判别结果进行评价,并以得分图展示样本在不同PLS-DA模型潜变量空间的分布情况。结果采用NIR光谱建立的PLS-DA模型对校正集和验证集的分类准确率分别为100%和84.62%,采用MIR光谱建立的PLS-DA模型对校正集和验证集的分类准确率分别为100%和92.41%,将NIR和MIR光谱数据融合后建立的PLS-DA模型对校正集和验证集的分类准确率均为100%,并且数据融合后,样本在PLS-DA模型潜变量空间呈现明显的分类聚集现象。结论数据融合方式可以显著提高姜半夏红外光谱鉴别准确率。本文为鉴别姜半夏及其伪品提供了新的研究思路和解决方法,为保证姜半夏临床用药安全、有效提供依据。
Objective To establish a rapid identification method for differentiating the ginger-processed Pinelliaternata from the ginger-processed Pinellia Pedatisectaby data infusion with near-infrared(NIR)and mid-infrared(MIR)spectroscopy techniques.Methods 22 batches of ginger-processed Pinelliaternata and 14 batches of ginger-processed Pinellia Pedatisecta were collected,and NIR and MIRspectral data of these samples were acquired.Discriminant modelsfor the individual data and the fused data were developed with the partial least squares-discriminant analysis(PLS-DA)method,and were evaluated by the classification accuracy(ACC).Results The ACC of PLS-DA model for NIR data was 100%in the training set and 84.62%in the test set respectively.The ACC of PLS-DA model for MIR data was 100%in the training set and 92.41%in the test set respectively.By combining NIR and MIR spectroscopy data,the ACC of PLS-DA model was 100%in both the training set and test set;moreover,after data fusion,the samples presented obvious clustering phenomenon in the latent variable space of PLS-DA model.Conclusion Such a data fusion approach might significantly improve the accuracy of infrared spectrum identification of ginger-processed Pinelliaternata and possibly provides a new research idea for identifying the ginger-processed Pinelliaternata to ensure its safety and efficacy.
作者
孙飞
陈雨
王凯洋
邱蕴绮
王淑美
梁生旺
Sun Fei;Chen Yu;Wang Kaiyang;Qiu Yunqi;Wang Shumei;Liang Shengwang(School of Traditional Chinese Medicine,Guangdong Pharmaceutical University,Guangzhou 510006,China;Key Laboratory of Digital Quality Evaluation of Chinese MateriaMedica of National Administration of Traditional Chinese Mediine,Guangdong 510006,China;Traditional Chinese Medicine Quality Engineering Technology Research Center of Guangdong Academies,Guangdong 510006,China;Guangdong Institute for Drug Control,Guangdong 5101802,China)
出处
《北京中医药大学学报》
CAS
CSCD
北大核心
2019年第10期862-868,共7页
Journal of Beijing University of Traditional Chinese Medicine
基金
广东省中医药局项目(No.20191193)
广东省医学科学基金项目(No.A2018295)~~
关键词
近红外光谱
中红外光谱
数据融合
姜半夏
质量控制
near infrared spectroscopy
mid-infrared spectrosopy
data fusion
ginger-processed Pinelliaternata
quality control
