Jiakai Lu

Accurately modeling the degradation of food antioxidants in oils is essential for understanding oxidative stability and improving food shelf life. This study presents an innovative machine learning approach integrating neural differential equations and sparse symbolic regression to derive a parsimonious differential equation for myricetin degradation in stripped soybean oil. Despite being trained on a small experimental dataset, the model successfully predicts degradation trends across a wide range of initial concentrations and extrapolates beyond the learning data. This capability demonstrates the robustness of machine learning for uncovering governing equations in complex food systems, particularly when experimental data is scarce. Our findings provide a framework for improving antioxidant efficiency in food formulations.

Due to the lack of a pathogen-killing process, foodborne outbreaks from contaminated fresh produce have been increasing worldwide. Hence, it is increasingly recognized that the washing step with sanitizers is important to control microbial contamination. Ozonated water is suggested as a substitute for chlorine-based sanitizers, addressing concerns about the effectiveness and environmental impact of chlorine-based sanitizers. However, using ozone as a sanitizer in the fresh produce washing process is still challenging because of its unstable and inconsistent antimicrobial effectiveness under various testing conditions. A meta-analysis was focused on the comparison of antimicrobial effectiveness between different ozonation methods commonly adopted in laboratory settings, including stationary pre-ozonated water, agitated pre-ozonated water, and sparging. The meta-analysis showed that the sparging method results in the highest microbial log reduction compared to other methods. We further developed meta-regression models based on three ozonation methods to identify key processing variables influencing the antimicrobial effectiveness of ozonated water. Attempts were made to link key processing variables to ozone stability and the mass transport phenomena involved in the washing process. This research will contribute to designing and developing a washing process to increase fresh produce safety by identifying key factors in each ozonation method and facilitate interlaboratory comparison studies.

Foam mat drying is an efficient and cost-effective method for producing high-quality strawberry powder.

This study investigated the efficacy of low-dose ozone microbubble solution and conventional aqueous ozone as inactivation agents against Tulane virus samples in water over a short period of time. Noroviruses are the primary cause of foodborne illnesses in the US, and the development of effective inactivation agents is crucial. Ozone has a high oxidizing ability and naturally decomposes to oxygen, but it has limitations due to its low dissolution rate, solubility, and stability. Ozone microbubbles have been promising in enhancing inactivation, but little research has been done on their efficacy against noroviruses. The study examined the influence of the dissolved ozone concentration, inactivation duration, and presence of organic matter during inactivation. The results showed that ozone microbubbles had a longer half-life (14 ± 0.81 min) than aqueous ozone (3 ± 0.35 min). After 2, 10, and 20 min postgeneration, the ozone concentration of microbubbles naturally decreased from 4 ppm to 3.2 ± 0.2, 2.26 ± 0.19, and 1.49 ± 0.23 ppm and resulted in 1.43 ± 0.44, 0.88 ± 0.5, and 0.68 ± 0.53 log₁₀ viral reductions, respectively, while the ozone concentration of aqueous ozone decreased from 4 ppm to 2.52 ± 0.07, 0.43 ± 0.05, and 0.09 ± 0.01 ppm and produced 0.8 ± 0.28, 0.29 ± 0.41, and 0.16 ± 0.21 log₁₀ reductions against Tulane virus, respectively (p = 0.0526), suggesting that structuring of ozone in the bubbles over the applied treatment conditions did not have a significant effect, though future study with continuous generation of ozone microbubbles is needed.

Concerns exist regarding the negative environmental impact and health risks associated with ocean fishing and aquaculture, such as stock depletion, pollution, biodiversity loss, and toxin presence. To address these concerns, plant-based seafood analogs are being developed. Our previous study successfully created plant-based scallop analogs using pea proteins and citrus pectin, resembling real scallops in appearance and texture. This study focuses on comparing the digestive fate of these analogs to real scallops, as it can impact their nutritional properties. Using an in vitro digestion model (INFOGEST), we simulated oral, gastric, and small intestinal conditions. The analysis revealed differences in the microstructure, physicochemical properties, and protein digestibility between the plant-based scallops and real scallops. The particle size and charge followed the following similar trends for both types of scallops: the particle size decreased from the mouth to the stomach to the small intestine; the particles were negative in the mouth, positive in the stomach, and negative in the small intestine. The protein digestibility of the plant-based scallops was considerably lower than that of real scallops. For instance, around 18.8% and 61.4% of protein was digested in the stomach and small intestine phases for the real scallop (80.2% total digestion), whereas around 8.7% and 47.7% of the protein was digested for the plant-based scallop (56.4% total digestion). The lower digestibility of the plant-based scallops may have been due to differences in the protein structure, the presence of dietary fibers (pectin), or antinutritional factors in the plant proteins. These findings are crucial for developing more sustainable next-generation plant-based seafood analogs.