Identification of Radix Bupleuri From Different Geographic Origins Using Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry and Support Vector Machine Algorithm

Abstract Background The geographic origin of Radix bupleuri is an important factor affecting its efficacy, which needs to be effectively identified. Objective The goal is to enrich and develop the intelligent recognition technology applicable to the identification of the origin of traditional Chinese medicine. Method This article establishes an identification method of Radix bupleuri geographic origin based on matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and support vector machine (SVM) algorithm. The Euclidean distance method is used to measure the similarity between Radix bupleuri samples, and the quality control chart method is applied to quantitatively describe their quality fluctuation. Results It is found that the samples from the same origin are relatively similar and mainly fluctuate within the control limit, but the fluctuation range is large, and it is impossible to distinguish the samples from different origins. The SVM algorithm can effectively eliminate the impact of intensity fluctuations and huge data dimensions by combining the normalization of MALDI-TOF MS data and the dimensionality reduction of principal components, and finally achieve efficient identification of the origin of Radix bupleuri, with an average recognition rate of 98.5%. Conclusions This newly established approach for identification of the geographic origin of Radix bupleuri has been realized, and it has the advantages of objectivity and intelligence, which can be used as a reference for other medical and food-related research. Highlights A new intelligent recognition method of medicinal material origin based on MALDI-TOF MS and SVM has been established.