Geographical Origin Identification for Tetrastigma Hemsleyanum Based on High Performance Liquid Chromatographic Fingerprint

Rare Chinese herbal medicine, Tetrastigma Hemsleyanum Diels et Gilg(T. hemsleyanum), has good medicinal value. However, the quality of T. hemsleyanum varies greatly from different origins. In order to prevent T. hemsleyanum from being inferior, the origin identification is particularly important. In this paper, the feasibility of a rapid identification of different origins for T. hemsleyanum using High performance liquid chromatographic (HPLC) fingerprint system was investigated. Twenty-three chromatographic peaks were identified as characteristic fingerprint peaks of T. hemsleyanum. Then, combining principal component analysis (PCA) and Adaboost method with logistic regression (Adaboost -LR) as weak classifier, the origin identification model is established. Experiments show that the correct rate of origin identification can reach 100% with the combination of HPLC fingerprint and the proposed identification model. Hence, HPLC fingerprint with Adaboost-LR is effective and feasible to discriminate the geographical origin of T. hemsleyanum.