Discrimination of raw and sulfur-fumigated ginseng based on Fourier transform infrared spectroscopy coupled with chemometrics

Ginseng (Panax ginseng), as a tonic and functional food in many countries and regions for thousands of years, is often sulfur-fumigated (SF) for storage and protection. However, our previous study indicated sulfur-fumigation could transform ginsenosides, the active components of ginseng, into sulfur-containing derivatives and thus affect the quality and safety of ginseng. In this study, a rapid and efficient method in discrimination of nonfumigated (NF) and SF ginseng was developed using Fourier transform infrared (FT-IR) spectroscopy coupled with multivariate statistical analysis. A total of 240 batches of raw spectra were obtained from NF and SF ginseng by FT-IR spectroscopy. After excluding the outliers, the different performance of 3 spectral signal enhancing methods, 3 modeling evaluation methods, and 4 model evaluation indexes were compared. The results demonstrated the feasibility of using FT-IR spectroscopy between 3650 and 3200 cm(-1) for the detection of sulfurfumigation in ginseng. After sulfur fumigation, the peak areas in fingerprint and functional group area varied significantly. In addition, the parameters of back propagation artificial neural network (BP-ANN) evaluation model are the highest, its accuracy = 91.67%, precision = 89.29%, recall = 96.15%, and F1 = 92.59%. The error rates of 3 models were k-nearest neighbor algorithm (KNN) (25.00%) > logistic regression (LR) (16.67%) > BPANN (8.33%). It can be concluded that FT-IR spectroscopy combined with multivariate statistical analysis has great potential in rapid discrimination of NF and SF ginseng, which can provide a valuable reference for the quality and effectiveness of edible and medicinal application of ginseng.