The growing food industry has led to the continuous expansion of the market of soy sauce and pot-roast meat products. However, traditional quality control methods have inherent limitations such as strong subjectivity, low efficiency, and poor predictability in areas including raw material selection, processing techniques, and flavor analysis, which severely restrict the high-quality development of the soy sauce and pot-roasted meat products industry. Machine learning, as an advanced data analysis and modeling technique, offers new solutions to these challenges. Against this background, this review discusses the application of machine learning in the quality control of soy sauce and pot-roast meat products, focusing on the assessment of raw meat freshness, analysis of processing suitability, selection and blending of spices, optimization of processing techniques, standardization of flavor prediction, quality grading based on data fusion, and shelf-life prediction. It also explores the current challenges and future trends in order to provide a technical reference for the quality control of soy sauce and pot-roast meat products.

This study aimed to elucidate the effects of four reheating methods (oven heating, boiling, microwaving, and air frying) and reheating cycles on warmed-over flavor (WOF) in prepared meat patties, prepared in our laboratory. Quality changes were investigated in multiple dimensions, including physicochemical properties, sensory characteristics, and volatile flavor compounds. The results indicated that WOF in the prepared patties was primarily characterized by rancid, sulfurous, and linseed oil-like odors. The key compounds contributing to WOF were identified as 2-hexanol, 3-octanol, 2,4-decadienal, 2-nonenal, hexyl acetate, and benzothiazole. Additionally, reheating cycles had an accumulative effect on the contents of these compounds, which significantly increases after more than two reheating cycles. Notably, microwaving and boiling were more likely to promote WOF formation, whereas air frying effectively suppressed the generation of key WOF-related compounds. Furthermore, air frying significantly reduced the hardness and chewiness of the patties (P < 0.05) and improved their color. In conclusion, these findings provide a theoretical foundation for both establishing flavor quality standards and optimizing quality attributes in processed meat products.

Freshness is not only an important factor affecting fresh meat quality classification, but also directly determines the shelf life of products. Therefore, it is important to construct a fresh meat freshness characterization system. However, the traditional methods are complicated to operate and time-consuming, so that they cannot give results rapidly and therefore cannot meet the current needs. Therefore, a fast and accurate technology for fresh meat freshness characterization is always urgently needed for the meat industry and is a research hotspot in academia. This paper summarizes the new technologies available for the determination of the freshness of fresh meat, including sensory bionics, intelligent response and spectrum analysis. It emphasizes on reviewing the advantages and disadvantages of these technologies and discusses future breakthrough directions. Lastly, this review is wrapped up with an outlook on future trends in this field, in order to provide a reference for further research.

Based on the highly positive correlation between chilled chicken freshness and the number of Pseudomonas, this study established a loop-mediated isothermal amplification (LAMP) method for rapid detection of Pseudomonas for the purpose of freshness detection. The results showed that the established LAMP system had high specificity for Pseudomonas, and the highest sensitivity was 3.05 × 10² CFU/mL. According to the measured values of Pseudomonas count, total bacterial count, total volatile basic nitrogen (TVB-N) content, pH, sensory properties of chilled chicken during storage, it was found that the total bacterial count and odor properties were the most accurate indicators of freshness, and the correlation between Pseudomonas count and them was the strongest. Pseudomonas count in the early stage of spoilage could be accurately identified by LAMP, and based on it the freshness of chilled chicken could be inferred. The method established in this study is time-saving, flexible, and applicable to a variety of scenarios, and it deserves to be popularized.

In this study, the physiology, biochemistry, safety and functional properties of Staphylococcus used as a starter culture were analyzed. Among the 33 strains of Staphylococcus, 11 tested positive for catalase and negative for hydrogen sulfide and their safety was evaluated by hemolysis, plasma coagulase and heat-resistant nuclease tests. In the functional study, 8 of the 11 strains were able to produce lipase and decompose fat and had high cholesterol degradation capacity. In the acid tolerance test, 4 strains were selected for strong acid resistance. The 4 strains could degrade casein, myofibrillar protein, sodium nitrite and some bioamines. They were identified as S. warneri 5F’-2, S. vitulinus 8A-1, S. warneri 5F-2 and S. succinus A31. These strains were positive for catalase, unable to produce hydrogen sulfide, and negative for hemolysis, plasma coagulase negative and heat-resistant nuclease. They had the functional properties of lipase production, cholesterol degradation, acid tolerance, protease production, myofibrillar protein degradation, nitrite degradation, and inability to degrade amino acids to produce bioamines, and biogenic amine degradation. Therefore, they had excellent fermentation potential, and could be used as strains for meat product fermentation.

A machine vision system was used to collect 948 three-dimensional images of chicken carcasses on the broiler slaughter line. This study aimed to develop a rapid method for the identification of primary dermatitis in chicken carcasses. The acquired images were preprocessed and segmented into 128 × 128 pixel pictures with grids. A total of 762 pictures of dermatitic skin and 775 pictures of normal skin were selected. A total of 24 feature values were extracted including third-order color moments, mean and variance of gray-level co-occurrence matrix features, Tamura texture features from the 1537 pictures and the segmentation threshold and area of dermatitis region. Based on dimensionality reduction by principal component analysis (PCA), linear discriminant analysis model, quadratic discriminant analysis model, support vector machine, random forest, back propagation neural network (BPNN) and GoogLeNet models were established, and their classification performances were compared. Among these models, the GoogLeNet model was the most effective in classifying dermatitic skin samples with an overall accuracy of 90.5% and an average detection speed of 122.65 sheets per second. The prediction accuracy of the model for chicken carcasses with dermatitis was 100%, while that for qualified chicken carcasses was 90%.

We undertook this study in order to explore the impact of stationary magnetic field (SMF) and alternating magnetic field (AMF) treatments at different intensities (1.5 and 4.5 mT) combined with freezing and/or thawing on the water retention and quality of precooked meat patties. The results showed that magnetic field-assisted freezing shortened the time required to pass through the maximum ice crystal formation zone, thereby accelerating the freezing rate. In addition, it reduced the thawing rate, thawing loss and cooking loss, decreased the proportion of free water, increased the proportion of immobilized water, and improved the hardness, elasticity, cohesiveness and chewiness of meat patties (P < 0.05). The 4.5 mT AMF treatment shortened the time required to pass through the maximum ice crystal formation zone by 417 min compared with the control group, and thawing loss and cooking loss decreased by 59.62% and 58.07% in the AMF-assisted freezing-thawing group, respectively. Furthermore, the hardness, elasticity, cohesiveness and resilience were significantly higher than those of the control group (P < 0.01). In summary, magnetic field-assisted processing holds potential in the freezing and thawing of prepared meat products.

Aeromonas spp. are commonly found in spoilage of chilled meat. Aeromonas salmonicida NCM 29 and A. salmonicida NCM 57 have been discovered the spoilage heterogeneity in degrading myofibrillar protein. In this study, the two strains were tested to uncover the discrepancy of meat spoilage in collagen-rich chilled meat and extracted collagen. The results revealed that chicken claws, riched in collagen, inoculated with NCM 29 showed higher values of total viable counts (TVC), total volatile basic nitrogen (TVB-N), pH, adhered cells, trichloroacetic acid (TCA)-soluble peptides, and protease activity compared to those inoculated with NCM 57. Furthermore, NCM 29 generated higher quantity of volatile organic compounds related to meat spoilage. The collagenase ((hemagglutinin protease (Hap)) secreted by NCM 29 has been identified as the key factor responsible for the observed discrepancies in spoilage, which gradually degraded collagen into peptides and hydroxyproline. The capacity of Hap to degrade type Ⅰ collagen in vitro indicated that it has apparent proteolytic activity, which could reduce the average particle size and alter secondary structure of collagen. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) further confirmed the degradation of the β-chain in collagen. These findings not only provide a theoretical basis for in-depth investigation of the meat spoilage mechanisms of Aeromonas spp., but also encourage us to take measures to avoid the spoilage of related bacteria such as Aeromonas spp. during the preservation process.

Chilled chicken is inevitably contaminated by microorganisms during slaughtering and processing, resulting in spoilage. Cutting parts of chilled chicken, especially wings, feet, and other skin-on products, are abundant in collagen, which may be the primary target for degradation by spoilage microorganisms. In this work, a total of 17 isolates of spoilage bacteria that could secrete both collagenase and lipase were determined by raw-chicken juice agar (RJA) method, and the results showed that 7 strains of Serratia, Aeromonas, and Pseudomonas could significantly decompose the collagen ingredients. The gelatin zymography showed that Serratia liquefaciens (F5) and Pseudomonas saponiphila (G7) had apparent degradation bands around 50 kDa, and Aeromonas veronii (G8) and Aeromonas salmonicida (H8) had a band around 65 and 95 kDa, respectively. The lipase and collagenase activities were detected isolate-by-isolate, with F5 showing the highest collagenase activity. For spoilage ability on meat in situ, F5 performed strongest in spoilage ability, indicated by the total viable counts, total volatile basic nitrogen content, sensory scores, lipase, and collagenase activity. This study provides a theoretical basis for spoilage heterogeneity of strains with high-producing collagenase in meat.

The aim of this study was to evaluate the factors influencing the inactivation effect of intense pulsed light (IPL) on Aeromonas salmonicida grown on chicken meat and skin, and to further develop prediction models of inactivation. In this work, chicken meat and skin inoculated with meat-borne A. salmonicida isolates were subjected to IPL treatments under different conditions. The results showed that IPL had obvious bactericidal effect in the chicken skin and thickness groups when the treatment voltage and time were 7 V combined with 5 s. In addition, the lethality curves of A. salmonicida were fitted under IPL conditions of 3.5–7.5 V. The comparison of statistical parameters revealed that the Weibull model could best fit the mortality curves and could accurately predict the mortality dynamic of A. salmonicida grown on chicken skin. And further a secondary model between the scale factor b and the treatment voltage in Weibull model was established using linear equations, which determined that the secondary model could accurately predict the inactivation of A. salmonicida. This study provides a theoretical basis for future prediction models of Aeromonas, and also provides new ideas for sterilization approaches of meat-borne Aeromonas.