| The traditional Chinese medicine (TCM) fingerprint technique is an important method for the quality control of TCM, which can integrally reflect the complex composition of TCM. The application of analytical instruments and chemometrics in fingerprints can provide the information of the origin, species and processed methods of TCMs, which would be used to guide the quality control of medicine. This is the research objective of the thesis. The significance of the thesis is the combination of fingerprint and chemometrics can achieve the quality control of TCMs even if the chemical ingredients in TCMs are uncentain. The similarity of chromatography obtained by same conditions can reflect the relationship among different ingredients, which can be used to judge the origin, processed methods and sources of TCMs and also can provide some reference for the quality control and clinical usage. However, the insufficient of fingerprint used for the quality control of TCMs lies in the uncertainty of the pharmacological action shown by one or some peaks. The correlation between fingerprint and pharmacological action of TCM is still fuzzy, which also is the problem to be resolved of fingerprint used for quality control of TCMs. In this thesis, near infrared spectroscopy and high performance liquid chromatography with the aid of chemometrics were applied to establish the fingerprints of some TCMs (including rhubarb and propossed rhubarb, Indigowoad Root, geo-herbalism medicine guang pericapium citri reticulate and pericapium citri reticulate, pericapium citri reticulate from pericapium citri reticulate viride). The information about processed methods and species can be obtained from the established fingerprints with the aid of chemometrics, therefore, the quality of the TCMs can be controlled effectively.First of all, the progress and insufficiency of TCM and the development of TCM fingerprint were mainly introduced. The application of chemometrics in TCM fingerprint and quality control was discussed in detail. Secondly, chromatographic fingerprints of raw, wine-treated and braised rhubarb samples were obtained by liquid chromatography with a diode array detector (LC-DAD). Eight LC peaks were selected for fingerprint analysis, and six compounds were identified. The analytical results resolved by principal component analysis (PCA) and supervised pattern recognition methods suggested that there were differences among these three kinds of rhubarb. Then, three, three kinds of Indigowoad Root samples, Radix Isatidis (RI), Rhizoma et Radix Baphicacanthis Cusia (RRBC) and simulated adulterated samples (AD) were discriminated by use of near infrared spectroscopy (NIR) and chemometrics. Genetic algorithm-partial least squares (GA-PLS), successive projections algorithm (SPA), and wavelet transform (WT) were employed to select the effective wavelengths, and then the selected wavelengths were used as the input variables for supervised pattern recognition methods. All pattern recognition models with WT pretreatment performed much better. In view of the quality control concerns, the NIR data of the Indigowoad Root samples were also used for the assessment of the quantitative composition of the samples with the aid of back propagation artificial neural networks (BP-ANN), and the obtained results were acceptable. After that, chromatographic fingerprint of geo-herbalism medicine guang pericapium citri reticulate and pericapium citri reticulate were established. Pericapium citri reticulate included guang pericapium citri reticulate and pericapium citri reticulate, and their appearance was very similar, but there were some differences in price and therapeutic effect. Superivised pattern recognition methods were used to establish the prediction pattern of guang pericapium citri reticulate and pericapium citri reticulate, and the obtained results were satisfactory. Finally, near infrared spectroscopy and high performance liquid chromatography with the aid of chemometrics were used to discriminate pericapium citri reticulate from pericapium citri reticulate viride. The effective wavelengths selected by successive projections algorithm (SPA) were used as the input data of supervised pattern recognition methods, but both results were poor. However, when the SPXY (sample set partitioning based on joint x-y distance) method was used to select and optimize the samples in the training set and the prediction set, the results from supervised pattern recognition methods could be well obtained. Fingerprint is an effective method for the quality control of TCMs, because it can not only comprehensively reflect the internal chemical ingredients in TCMs, but also can show the complexity between them. In this thesis, the chromatography of rhubarb and propossed rhubarb, Indigowoad Root, geo-herbalism medicine guang pericapium citri reticulate and pericapium citri reticulate, pericapium citri reticulate from pericapium citri reticulate viride were built and the efficient information was extracted from the chromatography, which was used for the further analysis with the aid of chemometrics. The result of the pattern recognition was consistent with the expectable experimental purpose. The result suggested that the method presented in the thesis could recognize the information on processed method, species and source of TCMs effectively and could supply approach for the quality control of TCMs. |
PDF Full Text Request