Rapid quantitative assessment of lipid oxidation in a rapeseed oil-in-water (o/w) emulsion by three-dimensional fluorescence spectroscopy

Oxidation status is one of the main quality characteristics for emulsion processing and consumption. In this study, three-dimensional fluorescence spectroscopy was developed to rapidly determine oxidation status in an oil-in-water (o/w) emulsion, the estimation of the storage time of o/w emulsion, and the discrimination of the o/w emulsion with different storage temperatures. Low-field nuclear magnetic resonance (NMR) was aimed at investigating the water and oil distributions in o/w emulsion. The digital data representing the fluorescence spectra and reference measurements of the o/w emulsions were analyzed using support vector regression (SVR) coupled with genetic algorithm (GA-SVR), search method (GRID-SVR), and particle swarm optimization (PSO-SVR). Under the optimum condition, the GA-SVR model shows the best correlation coefficients equal to 0.97 and 0.96 for the calibration set and prediction set, respectively. The overall results suggest that this system is useful for the rapid evaluation of emulsion quality.