Growth simulation of Pseudomonas fluorescens in pork using hyperspectral imaging

Abstract

Hyperspectral imaging (HSI) is a promising technique that can replace the time-consuming and destructive plate count method. In this study, growth models of Pseudomonas fluorescens (P. fluorescens) in pork using the plate count method and HSI were established. Hyperspectral information extracted by three different methods was used for the growth simulation of P. fluorescens. The Baranyi model in combination with the Ratkowsky square-root model (BRSR model) and the Huang model in combination with the Ratkowsky square-root model (HRSR model) for P. fluorescens in pork were directly constructed using a one-step method. Results showed that the correlation coefficients between the growth models established using HSI and the plate count method were high, ranging from 0.952–0.999. The adjusted coefficient of determination (Ra2) of simulation models reached 0.974, and the growth parameters obtained were consistent with those obtained using the plate count method. This study demonstrated that HSI may be potentially used as a new predictive microbial modeling technique in the future.