Identification of different degrees of processed Ginger using GC-IMS combined with machine learning

Summary: This study investigated the feasibility of using hyperspectral imaging and chemometric modeling to rapidly predict the ratio of gingerols to shogaols in dried ginger powder. The best-performing models used partial least squares regression and least absolute shrinkage and selection operator, with multiplicative scatter correction and second derivative Savitzky-Golay pre-processing.

Summary: This paper mainly explored the feasibility of rapidly determining the quality of dried ginger (DG) and its processed products using benchtop-NIR spectrometer and Micro-NIR spectrometer. Different models were established and analyzed qualitatively and quantitatively. The results showed that the Micro-NIR spectrometer combined with an appropriate wavelength selection method had better prediction ability. Therefore, it can accurately distinguish DG, LCG, MCG and HCG from each other and determine the gingerol compounds in DG and its processed products accurately.