基于三维指纹图谱的中药质控分析新方法研究

项目名称： 基于三维指纹图谱的中药质控分析新方法研究

项目编号： No.21275067

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 数理科学和化学

项目作者： 翟红林

作者单位： 兰州大学

项目金额： 65万元

中文摘要： 三维指纹图谱因其可全面反映样品的信息而获得日益广泛应用。本项目拟以中药色谱指纹三维图谱为研究对象，将其转换为数字图像后，利用小波分析、数字图像处理等技术优势应对实验测定图谱中存在的噪音信号及峰与基线的漂移等问题；挖掘能够表征样本三维指纹图谱之图像中多分辨信息的组成特征，剖析其对指纹图像的刻画性能；探寻能够稳健表征指纹图像中目标组分之特征，建立相应的模型以实现对不同来源样本的相似性等分析及多目标组分同时定量分析的新方法，并应用于黄连、血塞通等中药（剂）样品的质量分析。本立项的研究成果不但有利于降低复杂样本分析中的难度与成本、易于实现在线分析，为中药质控与评价提供更加客观、快速、准确的新方法，而且还可为探索其它三维图谱（如三维荧光、LC-MS等）的分析与应用提供有益的启示。

中文关键词： 化学三维图谱；定性定量分析；化学计量学；数字图像处理；

英文摘要： Three-dimensional (3D) spectra have been widely applied in analytical chemistry owing to their comprehensive representation for the characteristics of samples. In this study, based on the digital grayscale image convered from 3D fingerprint spectra that were obtained by means of high performance liquid chromatography (HPLC) coupled with diode array detector (DAD), the qualitative and quantitative analyses for several productions of traditional Chinese medicines (TCM) will be investigated. The noise signals and shifts of peaks and baseline will be corrected using wavelet analysis and the technology of digital image processing. We will try to mine the image's features that can describe the composition of information with multi levels in the image of 3D HPLC/DAD fingerprint spectrum, and evaluate their performance. With the robust features of image corresponding to the compounds, the approach to the qualitative analysis of samples obtained from different sources will be developed, and the novel method of simultaneous quantitative analysis for the multi target compounds in sample will be explored and established. Furthermore, our methods will be applied on the qualitative and quantitative analyses of several TCM productions such as Huanglian and Xuesaitong injection. Our study will develop the faster, accurate and

英文关键词： Chemical three-dimensional spectra；Quantitative and qualitative analyses；Chmometrics；Digital image processing techniques；