European Food Research and Technology最新文献

Food safety is a critical public health issue requiring a shift from reactive crisis management to proactive, predictive governance. This study introduces an end-to-end deep learning framework, built upon a recurrent neural network architecture, to forecast monthly dairy-related food safety notifications from the EU’s Rapid Alert System for Food and Feed (RASFF). By integrating a diverse set of economic, environmental, social, and technological indicators from five EU countries (2001–2024), the model provides explicit, forward-looking risk predictions. The framework’s core innovations include a robust in-model strategy for handling missing data, which avoids traditional imputation bias, and an interpretability method that reveals the direction and temporal scale of risk drivers. The model achieved strong predictive performance, strategically prioritizing public health sensitivity with a high recall of 80.34%, a balanced F1-score of 71.21%, and an AUC-ROC of 0.8292. The interpretability analysis revealed that systemic risk is driven by a composite of factors with distinct temporal signatures: chronic, year-round pressures (e.g., antibiotic usage), seasonal risks (e.g., temperature), and long-lagged structural influences (e.g., agricultural income). By unifying forecasting with multi-level interpretation, this research provides a validated blueprint for intelligent early warning systems. It offers actionable, temporally-specific insights that enable policymakers, industry, and researchers to advance a more agile and data-informed approach to food safety.

Food quality control is crucial in the agricultural sector, especially in the case of table olives intended for fresh consumption, whereby selection and classification of drupes are important in post-harvest to contain product losses and in pre- and post-processing to attract consumer attention. In recent years, autonomous automatic inspection tools based on advanced technologies, i.e. artificial intelligence (AI)-powered image analysis, have been developed to objectively and non-invasively determine certain optical products characteristics related to quality parameters. This study proposes a real-time quality classification process of freshly harvested table olives belonging to three different Italian cultivars, both in terms of ripening stage and presence/absence of external and internal defects, using a mechano-optoelectronic system equipped with a VIS-NIR hyperspectral camera and an RGB camera, enhanced with AI algorithms. A shallow neural network model was developed for VIS-NIR images to identify cultivars, ripening degree and internal defects of olives, achieving better performance (99% accuracy) for cultivar classification. RGB images were processed with You Only Look Once (YOLO) Convolutional Neural Networks to discriminate regular from damaged olives, both by single cultivar and all cultivars together. The model considering all cultivars together showed good performance in detecting damaged olives, with a mean test accuracy of 0.87. The real-time mechano-optoelectronic system used to analyze olives through RGB and VIS-NIR images, in combination with AI algorithms, proved to be a good tool to qualitatively classify olives, regardless of their maturity and cultivar. Through appropriate scale-up the system could be used and implemented in processing lines for industrial purposes, making the agricultural system more effective and efficient and achieving higher quality production.

Despite the fact that Pseudomonas aeruginosa is a well-known opportunist pathogen, it remains an understudied microorganism in the context of food safety. Due to its high adaptation capability, rapid reproduction and minimal growth necessities, P. aeruginosa is a common bacterium in the food chain. The resistance of P. aeruginosa to various antibiotics in clinical isolates has been extensively investigated and in-depth characterized. Conversely, information on food-borne isolates of this genus and species is insufficient. In this study, a total of 208 colonies were isolated from 144 food samples collected from different provinces and subjected to morphological, physiological, and chemical analyses. Based on the results, total 22 isolates obtained from milk (12) and chicken (10) samples and exhibiting characteristics specific to P. aeruginosa were identified at the molecular level using the 16 S rRNA region. The antibiotic resistance profiles of the identified isolates were evaluated against azithromycin and erythromycin (macrolides), cefoperazone and ceftazidime (cephalosporin), meropenem and imipenem (carbapenem), as well as amikacin (aminoglycoside) among others. As a result of 16 S rRNA analysis, it was determined that only the 10 isolates obtained from raw milk samples were P. aeruginosa. The prevalence of antibiotic resistance among P. aeruginosa isolates was found to be variable. All isolates were observed to be susceptible to amikacin, imipenem, and piperacillin/tazobactam, while they were resistant to erythromycin, nalidixic acid, ampicillin, kanamycin, and chloramphenicol. Only the RM16 strain was susceptible to tetracycline, RM18 to meropenem, and RM20 and RM24 showed intermediate susceptibility to gentamicin.

This study explores the development of a sustainable, gluten-free beer using regionally sourced ingredients, including rice, potato peels, and native forest botanicals (St. John’s Wort, Juniper, and Helichrysum). The research aimed to optimize the fermentation of rice-based wort at lab-scale by incorporating potato peels and various amylolytic enzymes to improve brewing efficiency and sugar profile complexity. Rice malt was produced at the experimental facilities of KU Leuven (Belgium), and fermentation trials were conducted using both malted and unmalted rice. The study compared traditional decoction mashing, a novel thermostable enzyme (Enzyme X), and commercial enzyme preparations. Decoction-based trials exhibited limited fermentation efficiency (RDF ~ 45%), whereas Enzyme X significantly enhanced starch hydrolysis and fermentation performance (RDF 78–81%). The combination of unmalted rice with Enzyme X yielded the highest efficiency, reaching RDF values of up to 86%. Results demonstrated that the novel enzyme significantly enhanced starch hydrolysis, reducing mash duration by approximately 30 min while improving overall fermentation efficiency. Additionally, the integration of potato peels (10–20%) contributed to higher RDF values (up to 86%) and shifted the sugar profile toward greater maltose and oligosaccharide content, mitigating glucose dominance and potentially influencing yeast metabolism and flavor profile. These findings suggest that enzymatic innovations and agro-industrial byproducts, such as potato peels, could play a key role in advancing sustainable brewing practices while maintaining desirable fermentation characteristics in gluten-free beer production.

Strawberry yield and quality in semi-arid regions depend critically on genotype selection and propagation method. We evaluated six cultivars (Rubygem, Amiga, Sweet Ann, Kabarla, Festival, and Fortuna) propagated via frigo and fresh seedlings under open-field conditions in Diyarbakır, Turkey, across two consecutive years to determine optimal cultivar-propagation combinations for challenging environmental conditions. Genotype and seedling type significantly influenced yield, color parameters, soluble solids, phenolic compounds, organic acids, sugars, and ascorbic acid (p ≤ 0.05). Frigo seedlings ensured earlier flowering and 123% higher yields (216.84 vs. 97.13 g/plant), with Rubygem achieving maximum productivity at 331.72 g/plant. However, fresh seedlings produced 11% brighter fruit with 23% higher pelargonidin content (38.4 vs. 31.2 ppm), demonstrating a yield-color quality trade-off. Soluble solids remained unaffected by propagation method (10.7°Brix), with Amiga and Rubygem showing highest levels (12.1 and 11.6°Brix). Phenolic profiles were cultivar-dependent: Festival and Amiga dominated in quercetin (42.1 and 39.8 ppm), Fortuna in pelargonidin (41.8 ppm), and Kabarla in ellagic acid (4.30 ppm). Frigo plants accumulated 39% higher ellagic acid (2.71 vs. 1.95 ppm), suggesting enhanced stress-protective mechanisms. Festival exhibited highest citric acid (1.19%), while Kabarla showed superior succinic acid (0.58%). Total sugars were highest in Rubygem (9.11 g/100 g FW) and frigo Festival (10.20 g/100 g FW). Fresh Festival demonstrated exceptional ascorbic acid (77.6 ppm), 41% higher than frigo counterparts. Multivariate analysis revealed strong negative correlation between pelargonidin and yield (r = − 0.59), confirming resource allocation trade-offs. Rubygem, Amiga, and Sweet Ann emerged as optimal choices for semi-arid production, while Festival represents valuable germplasm for vitamin C enhancement and climate-resilient breeding programs.

Pectic oligosaccharides (POS) exhibit promising bioactive properties, primarily attributed to their simple structure, low viscosity, high solubility, and low molecular weight (Mw). In this study, a simple method based on the Fenton reaction and ethanol precipitation was used to prepare and isolate citrus pectic oligosaccharides (CPOS). The structural characteristics of the resulting CPOS were analyzed using FTIR, SEM, NMR, and UPLC-Q-TOF-MS, and their antioxidant activities were evaluated. Structural characterization revealed that the obtained CPOS consisted mainly of galacturonic acid (GalA). Further analysis via NMR and UPLC-Q-TOF-MS confirmed that CPOS with a degree of polymerization (DP) up to 13 was rich in GalA and galactose (Gal). Antioxidant activity assays demonstrated that CPOS exhibited strong DPPH (91.57% at 2 mg/mL) and ABTS (87.14% at 1 mg/mL) scavenging rates. This study indicated that a simple method based on Fenton reaction and ethanol precipitation can be used to produce CPOS with antioxidant activity, providing a technical reference for the industrialized low-cost production of POS and the development of functional food ingredients.

Euryale ferox seeds can be classified into three maturity stages—tender, green, and old—based on their growth period and suitability for consumption or processing. Maturity inevitably influences their physicochemical and functional properties. This study systematically examined the nutritional composition, microstructure, and functional properties of whole flour and isolated starch from seeds at different maturity stages. The results showed that increasing maturity significantly increased starch content (from 77.00% to 79.36%) while decreased amylose content (from 39.15% to 36.15%). Compared with tender seeds, the flour and starch from old seeds exhibited significantly higher relative crystallinity, pasting temperature, and enthalpy but significantly lower peak viscosity (965 cP) than those of tender (3624 cP) and green (2316 cP) seeds (p < 0.05). Rheological analysis revealed significant increases in G' and G'' values for E. ferox flour and starch with increasing maturity. These findings demonstrate that seed maturity significantly affects starch structure and functionality, particularly gelatinization and rheological behavior. Overall, this study provides a theoretical basis for the rational selection of E. ferox raw materials based on processing performance, thereby facilitating targeted improvements in end-product quality.

The valorization of food waste biomass for the extraction of bioactive compounds presents a sustainable solution to global food waste challenges while offering significant economic and environmental benefits. This review comprehensively examines advanced green extraction technologies such as ultrasound-assisted extraction (UAE), microwave-assisted extraction (MAE), supercritical fluid extraction (SFE), and enzyme-assisted extraction (EAE) for recovering polyphenols, carotenoids, flavonoids, and other high-value compounds from fruit, vegetable, cereal, and animal-derived waste. Highlighting optimized extraction parameters showed that modern techniques outperform conventional methods in yield, efficiency, and cost-effectiveness, with UAE reducing manufacturing costs by up to 84% compared to Soxhlet extraction. The review also addresses pretreatment strategies, purification innovations (e.g., flash chromatography), and emerging solvents like natural deep eutectic solvents (NADES) with lower environmental impact, higher efficiency, recyclability and biodegradability. Challenges such as contamination risks, temperature sensitivity, and scalability are critically analyzed, alongside future directions integrating nanotechnology, artificial intelligence, and hybrid extraction systems. By bridging gaps in waste classification, compound identification, and process optimization, this study underscores the potential of food waste as a resource for pharmaceuticals, nutraceuticals, and functional foods, aligning with circular economy goals and sustainable development.

The plastron of yellow pond turtles is a kind of high-value Chinese herbal medicine with abundant protein resources. Turtle plate has the potential of anti-inflammation, analgesia, enhancing immunity, nourishing yin and suppressing yang, osteoporosis and so on, and is often developed and utilized as a functional food. However, the cartilage protective activity of yellow pond turtle peptides (YPTP) in knee osteoarthritis (KOA) and the associated mechanism have not been reported. Thus, this study investigated the therapeutic potential of YPTP for KOA and HPLC-QTOF-MS identified the structure of peptides. The results revealed that YPTP exhibited significant therapeutic effects against KOA progression. It was found that COX-2 and iNOS were the primary target proteins of YPTP in KOA. Molecular docking analysis further validated the 5 different YPTP-derived peptides that interacted with COX-2 and iNOS via hydrogen bonding and hydrophobic interactions. Using an animal model, we revealed that YPTP significantly suppressed the contents of COX-2, iNOS, and MMP-3 while upregulating COL II to alleviate inflammation and minimize cartilage damage. However, reduced YPTP dose enhanced IL-1β contents in synovial membranes, which requires further exploration.

This study investigates the biochemical composition of six apricot cultivars (‘Precoce de Tyrinthe’, ‘Rubista’, ‘Domino’, ‘Flopria’, ‘Orange Ruby’, and ‘Çağatay Bey’) selected from commercial orchards in Mut district of Mersin, Turkey in 2024. Four sugars were identified: sucrose (3.925–5.864 mg/100 g), glucose (988.95–2.202 mg/100 g), xylose (79.76–222.11 mg/100 g), and fructose (446–1462 mg/100 g). Total sugar content ranged from 6.235 mg/100 g (Domino) to 9.424 mg/100 g (Çağatay Bey), with sucrose and glucose being predominant across all cultivars. Eight phenolic compounds were detected and quantified, with catechin (0.49–1.74 mg/100 g) and ellagic acid (0.61–3.62 mg/100 g) being most abundant. Significant cultivar variation was observed in phenolic profiles, with Precoce De Tyrinthe exhibiting the highest caffeic acid and catechin content, while Çağatay Bey demonstrated the highest ellagic acid concentration. These phenolic signatures provide biochemical markers for cultivar authentication and antioxidant capacity assessment. A comprehensive volatile analysis identified 53 compounds, including 12 alcohols, 11 aldehydes, 7 esters, 8 ketones, 3 acids, and 12 terpenes. Alcohols (24.28–72.89%) and aldehydes (11.55–48.55%) were the dominant volatile groups across all cultivars. Key aroma compounds included 1-hexanol (0.43–7.08%), (E)-2-hexen-1-ol (11.17–25.08%), linalool (6.36–54.86%), hexanal (3.65–18.52%), and benzaldehyde (0.54–18.46%). Cultivar-specific volatile profiles revealed distinct aroma signatures, with Flopria showing exceptionally high linalool content (54.86%) suitable for premium applications. Principal component analysis explained 94.82% of total variance across four components, enabling systematic cultivar discrimination based on integrated biochemical profiles. These comprehensive data provide quantitative biochemical fingerprints for cultivar characterization, breeding program optimization, and targeted market applications in international apricot production systems.