Application Of Electronic Nose In Quality Evaluation Of Chrysanthemum

Chinese Materia Medica (CMM) as the crystallization of human wisdom for many thousand years, makes a significant contribution to human. In modern society, CMM has been given more and more great importance. Since the clinic safety and efficacy are closely related to CMM quality, the quality evaluation of CMM is particularly important.Macroscopic identification is the traditional authentication method to identify Chinese herbal medicines (CHMs), which through observing, touching, smelling and tasting to look at its shape, color, odor, taste, etc. Morphological identification as a quality control method of CMM, not only reflects the clinical practice experience for thousands of years but also plays an important role in quality evaluation. However, macroscopic identification has a strong subjective and is not conducive to the promotion and implementation.With the progress of science and technology, means of modern science and technique, such as Thin Layer Chromatography (TLC), High Performance Liquid Chromatography (HPLC), Gas Chromatography (GC) and Gas Chromatography-Mass Spectrometry (GC-MS), have been already employed into CMM. However, its disadvantages, such as time-consuming, expensive, destroying samples, environmental pollution and lack of integrality cannot be ignored. Therefore, the macroscopic identification of CMM needs to be validated and standardized by modern scientific approaches.Electronic nose (E-nose), as a newly born odor identification technology, has been paid more and more attention. In the past decades, E-nose has been widely applied in the fields of environment, food, medicine and so on.In this paper, it’s planned to be implemented based on macroscopic identification and physicochemical test with the CMM Chrysanthemi Flos as an object. And the E-nose is the important key in this research in order to make macroscopic identification more digital and objective. First of all, the optimal experimental conditions of E-nose are explored. And then, the pattern recognition algorithms are the next key in this research which included DFA, PCA and 12 ways of machine learning (ML), et al. Secondly, optimized identification models are built up according to the results of DFA PCA, Bagging, BF Tree, J48, Random Tree and Random Forest. Finally, an optimization identification system of Chrysanthemi Flos is due to be built. Meanwhile, other methods of quality evaluation are combined with the result of E-nose to evaluate the quality more holistic.In this research, E-nose, as a simple, nondestructive, sensitive and rapid technology without organic reagents, could differentiate the different cultivars of Chrysanthemi Flos. What is more, some more innovative pattern recognition algorithms and the optimization of sensor arrays were combined with E-nose technology to broaden the thinking of E-nose’s data processing. In conclusion, E-nose, like the human olfactory system, can be used as a supporting technique for the quality evaluation of the CMM.