Non-destructive Assessment Of Microbial Contamination And Edible Safety In Chilled Pork Based On Optical Technique

Chilled pork is increasingly becoming the mainstream of raw meat consumption in China. It has great advantages of succulent fresh, high nutritional value, easy to split. However, the raw pork is a highly perishable food product and easily contaminated by microorganism that is able to cause food-borne disease. It has great practical significance for rapid, accurate and nondestructive detection of chilled pork microbial contamination edible safety. Recently, rapid, non-invasive and real-time optical technique for detection of agro-product quality and safety has great practical significance with its unique advantages. Therefore, in this study, the indexes for chilled pork edible safety and methods for rapid, nondestructive and real-time prediction of meat edible safety and microbial conmaination ware studied. The results and creations were listed as follows:(1) The key microbial indexes and and edible safety evaluation indexes of chilled pork were studied. The specific spoilage organism Pseudomonas spp. and total viable counts (TVC) were determinated as key microbial indexes. Microbiological, physicochemical and organoleptic characteristics such as the total viable counts (TVC), Pseudomonas spp., total volatile basic-nitrogen (TVB-N), pH value and color parameter L*were determined to appraise chilled pork edible safety in the cold chain.(2) The optical scattering characteristics of chilled pork samples from were extracted based on Lorentz function, Gompertz function and Boltzmann function. Lorentz function and Gompertz function were used to fit the scattering profiles of pork based on the partial least square non-linear fitting in the study, and the results show that the two functions can be used to extract the optical scattering profiles of pork effectively. Support vector machine (SVM) were applied as statistical tool to establish analytical method based on pork edible indexes and optical scattering spectra. The accuracy and stability of the prediction models were be improved by SVM based on Lorentz three parameters combination [abc].(3) Predictive models for specific spoilage bacteria and total viable counts (TVC) were established by SVM based on hyperspectral scattering technique and near infrared spectroscopy rapidly and non-invasively. For TVC and Pseudomonas spp., the SVM models with Lorentz parameters combination [abc] produced best prediction results with Rv of0.968and0.948, SEV of0.410and0.711log CFU/g, respectively. Furthermore, the edible safety of chilled pork samples based on two near infrared spectroscopy system which included400-1000nm Vis/NIR spectra and1000-2500nm NIR spectra was detected. Different pre-treatment methods and spectral variable screening methods were used to improve the model performance.(4) A rapid, nondestructive and accurate optical method was set up based on microbial, physicochemical and organoleptic indexes to appraise chilled pork edible safety comprehensively. All stored chilled meat samples were classified into three grades:"fresh","semi-fresh", and "spoiled". Bayesian classification model was superior to PLS-DA with overall classification accuracy of92.86%.