Application Of Chemometrics To The Analysis Of The Traditional Chinese Medicines And The Development Of Their Fingerprints

Traditional Chinese medicine(TCM) has become a thriving industry, not just in China, but worldwide. In many cases, toxicity or side effects associated with Western Medicine have worsened human diseases, which make people more inclined to try natural medicine, and traditional methods of treatment. In the new century, TCM will meet with great changes in both scientific development and social life. They will create both opportunities and challenges for TCM. Some experts have even forecast that the 21st century will be the era of TCM, with the modernization of TCM, it may become the cornerstone of Chinese industries. TCM is an analytical system with too much complexity. The pharmaceutical activity of TCM come from the interactional of all the chemical components in it. The explicit characterization of chemical substance has become the bottleneck in the process of TCM standardization and internationalization. There is two way to overcome the bottleneck, they would be discussed in details in the thesis. One side, analyze the detailed chemical components in the TCM, another side, develop the fingerprint of the TCM to meet the authentication and quality control of TCM.The essential oil of Schisandra chinensis(Turcz.)Baill. was analyzed with GC/MS, heuristic evolving latent projections(HELP) resolution and overall volume integration method. HELP method along with the data from GC/MS can conduct the peak purity examination and resolution of overlapping peaks to obtain pure chromatogram and MS spectrum of each component. 56 components are separated and 49 of them were qualitatively and quantitatively analyzed. All the components represent about 98.27% of the total content. The resolved pure chromatogram and MS spectrum can greatly enhance the reliability of similar searches in the commercial MS database and thus contribute to the accuracy of the qualitative and quantitative analysis of the essential oil.The two-dimensional data obtained from GC/MS is investigated to determine qualitatively and quantitatively chemical components of volatile factions inSchisandra chinensis (Turcz.) Baill. from six different extraction methods. Subwindow factor analysis (SFA) was specially used to confirm the same components determined from different samples. With the help of SFA and other chemometric techniques, detection of peak purity in chromatograms can be first addressed, and then overlapping peaks arc resolved into the pure chromatographic profile and mass spectrum of each component. The results obtained show that the accuracy of qualitative and quantitative analysis could be greatly enhanced by the chemometric resolution methods. The chemometric resolution techniques upon the two-dimensional data can be quite promising tools for the analysis of the complex samples like traditional Chinese medicines. It is also found that different extraction methods obtain different chemical components both in quality and quantity aspects for the volatile fraction of Schisandra chinensis (Turcz.)Baill..The GC/MS data from essential oil of Houttuynia cordata Thunb in Huaihua, Hunan province, were resolved with chemometrics resolution method and local smoothing for component identification method. 55 components were elucidated by qualitative and quantitative analysis. The results will be most important in thedevelopment of the fingerpprint for the Houttuynia injection.In order to improve detection ability and quality of resolution of overlapping peaks with low signal to noise ratio data obtained from GC-MS, the effect of heteroscedastic noises is investigated in the present paper. A new index named smoothing distortion is first developed for evaluating the smoothing efficiency. Roughness penalty smoothing method recently appearing in chemometrics is then compared with wavelet de-noising technique and convolution smoothing approach under condition of heteroscedastic noises. The performance of the methods is assessed using both simulated and experimental GC-MS data. The results obtained show that the roughness penalty method can not o...