The Research About Spectrum-effect Relationship Of Traditional Chinese Medicine

The development of Chinese medicine runs through the whole long history of Chinese civilization. The research of traditional Chinese medic-ine(TCM) needs to keep pace with the times and becomes internationaliz-zation and standardization gradually. So we need to establish a standardi-zed quality evaluation system which is widely accepted by the internatio-nal community. The appearance of fingerprinting brought a new direction to the further development of TCM. That how to combine fingerprinting of TCM and its clinical efficacy, find out the peaks closely related to the effectiveness and find out the really useful substance is a key and a hot spot about the research of the spectrum-effect relationship. The settlement of this problem can guide for the practical application and clinical medicine and provide a good way for the internationalization and modernization of Chinese medicine. The composition of TCM is compli-cated, even through single herbs, which contains huge ingredients, not to mention all kinds of compound compatibility of medicines whose compo-sition is more complicated. Facing with complex TCM, we can imagine the difficulties in establishing a unified, scientific quality standard and finding out the compounds associated with efficacy in numerous material. Many researchers used modern technology means and mathematical methods to study spectrum-effect relationship. They have tried various methods, including gray correlation degree, which can be used to determ-ine the material composition related to the efficacy. On this basis, They also obtained good results. In this paper, we give a new method to determine the weight on the basis of the existing correlation and apply it to the examples. The effect of our method is good and it can also find out the chromatographic peaks related to the efficacy.In the research about spectrum-effect relationship, setting up an internal quality evaluation system to accurately predict efficacy is also very important. Facing with such complex nonlinear relationship between fingerprinting and effect of Chinese medicine, the traditional multiple linear regression and multivariate statistical modeling method are difficult to evaluate the intrinsic quality of TCM, and can’t predict efficacy accura-tely. The artificial neural network has an advantage of solving nonlinear problems, so it has become the trend to use artificial neural network to predict efficacy in recent years. In the current research on neural network, the scholars mainly use BP neural network to predict efficacy, but the BP neural network is easy to fall into local minimum and its convergence speed is slow. The Generalized Regression Neural Network(GRNN) has strong nonlinear mapping ability. Based on this, the paper propose a method using GRNN to predict the efficacy about the obtained data of peaks.