Mycotoxin contamination in cereals poses a significant food safety issue. In this study, aflatoxin B1 (AFB1), fumonisin B1 (FB1) and ochratoxin A (OTA) were determined in 114 cereal samples from the Fuyang region of Anhui Province, China, and dietary risks were assessed. Among them, 83.8% maizes were AFB1-positive, 100% maizes and 6.5% wheats were FB1-positive. Measured mycotoxin migration rates from raw materials to processed foods were 55.79-77.0% for boiled food and 52.51-96.60% for non-boiled food; applying these rates as processing factors to intake estimates altered exposure calculations and thereby markedly affected the risk assessment. Ultimately, potential carcinogenic risks were observed for the groups of ‘all consumers’ and ‘children’. Although maize-based food comprised only 3.4% of cereal consumption, maize contributed disproportionately to estimated carcinogenic risk relative to wheat. The method of this study improves the accuracy of risk assessment and underscores the need to prioritize mycotoxin control in maize in the Fuyang region of Anhui Province.
Ensuring the safety of infant nutrition is a public health priority, particularly regarding oral exposure to contaminants. Uranium produces toxicities such as neurotoxicity, reproductive toxicity, liver toxicity, etc. upon exposure. This article reviews the occurrence of uranium in breast milk and infant formula based on 15 reports from 13 studies (published between 2011 and 2025) retrieved following a systematic search of databases PubMed and Scopus. Inductively coupled plasma mass spectrometry (ICP-MS) and microwave digestion were respectively the most commonly used determination and extraction methods in studies that reported uranium concentration in both matrices. Uranium was detected in five out of seven studies on breast milk, with concentrations ranging from 0.01 to 160.97µg/L and the highest level found in samples from Brazil. On the other hand, all eight studies on infant formula reported uranium contamination, with concentrations reaching 20µg/kg in a study done in Brazil. Based on the available literature, infant formula is potentially a greater source of uranium exposure than breast milk.
This study reports the development and national certification of neat-substance certified reference materials (CRMs) for the Aconitum alkaloids hypaconitine, mesaconitine, and yunaconitine. The certified materials were prepared from commercially sourced crude extracts of the corresponding alkaloids by preparative liquid chromatography followed by concentration and vacuum freeze-drying. Pre-certification purity was evaluated by mass balance and quantitative NMR, whereas final values were assigned through an eight-laboratory HPLC study with mass-balance correction. Certified purities were 99.36% (U = 0.25%, k = 2), 99.28% (U = 0.34%, k = 2), and 98.57% (U = 0.52%, k = 2), respectively. Identity was supported by UV, IR, HRMS, and one-dimensional NMR. Certificates specify a 1-mg minimum intake, −18°C storage, 2–8°C transport, and 2.5-year validity. The certification record includes 33 months of monitoring at 4°C. A positive mesaconitine trend at 25°C and degradation at 50°C support low-temperature handling. These CRMs are intended for calibration and analytical quality control; matrix-specific recovery and matrix effects require separate validation.
Forest cultivated dried ginseng is a valuable medicinal material whose cultivation age influences its quality and market value, yet conventional identification methods are often destructive or inefficient. In this study, hyperspectral imaging (HSI) combined with a Random Forest model optimized by Improved Holistic Swarm Optimization (IHSO-RF) was employed for the nondestructive discrimination of samples aged 5, 6, and 7 years. Among nine preprocessing methods, Savitzky-Golay smoothing combined with multiplicative scatter correction (SG+MSC) provided the best performance. CARS and RFE were applied to extract informative spectral wavelengths and textural features, respectively, for feature fusion. The proposed IHSO framework integrated cascaded Logistic Tent chaotic initialization, an adaptive polynomial exponential step size strategy, and adaptive Lévy flight with sinusoidal mutation to enhance population diversity, balance global exploration and local exploitation, and improve convergence performance. The Fusion-IHSO-RF model achieved the best classification performance, with accuracy, precision, recall, and F1-score values of 0.9489, 0.9610, 0.9622, and 0.9616, respectively. SHAP analysis indicated that spectral features contributed strongly to age discrimination, while texture features provided complementary information. This method enables accurate non-destructive classification of adjacent cultivation ages of forest-cultivated dried ginseng, providing an effective strategy for quality grading and market standardization of traditional Chinese medicinal materials.
Phthalate esters (PAEs) are ubiquitous plasticizers that can migrate from polymeric packaging into foodstuffs, particularly under elevated storage temperatures. This study quantified three priority PAEs DEHP, DBP, and BBP in thirty brands of packaged coffee marketed in Iran and characterized the associated non-carcinogenic and carcinogenic health risks using a 10,000-iteration Monte Carlo simulation. Each sample was analyzed in triplicate by LC-MS/MS. Mean concentrations were 4.19, 0.35, and 1.90µgkg⁻¹ for DEHP, DBP, and BBP, respectively, with DEHP the dominant congener. Follows the USEPA framework, estimated daily intake (EDI), hazard quotient (HQ), cumulative hazard index (HI), and incremental lifetime cancer risk (ILCR) were derived for five age classes (children, teenagers, youth, middle-aged, and adults) of both sexes. Across all ten subpopulations, the commodity-specific HI remained far below unity, ranging from 6.15×10⁻⁶ (male children) to 1.62×10⁻⁵ (female youth), with 95th percentile values ≤ 4.44×10⁻⁵, indicating negligible non-carcinogenic hazard. The DEHP ILCR was likewise far below the USEPA benchmark of 1×10⁻⁶, ranging from 1.63×10⁻¹⁰ (male children) to 1.17×10⁻⁹ (middle-aged females), with 95th percentile values ≤ 3.17×10⁻⁹, indicating negligible carcinogenic risk. Sensitivity analysis identified DEHP concentration (ρ = 0.76) and coffee ingestion rate (ρ = 0.57) as the principal drivers of exposure variance. Although the present exposure from coffee alone is acceptable, DEHP predominance warrants continued surveillance, improved packaging materials, and temperature-controlled storage.
This study examined the in vitro bioaccessibility fractions (%BF) of microelements (Cr, Cu, Fe, Mn, Ni, and Zn) in five species of medicinal and edible mushrooms from the Brazilian Atlantic Forest. We evaluated the %BF of these elements using a simulated gastrointestinal digestion protocol (INFOGEST 2.0) and microwave-induced plasma optical emission spectrometry (MIP-OES). Additionally, we examined their contribution in relation to the recommended daily intake (RDI). The %BF values ranged from 11.50% to 85.17%. Although Fe exhibited high total levels, it contributed only 1.3-3.0% of the RDI. In contrast, Mn and Cr exhibited high gastrointestinal release, providing up to 39.4% and 44.3% of the RDI, respectively. Overall, these results highlight the importance of incorporating in vitro bioaccessibility data into the multi-element characterization of mushrooms.
Soluble solids content (SSC) and firmness are key quality attributes for assessing the maturity, postharvest quality, and commercial value of Xinmei plum. This study developed a nondestructive analytical approach combining multidimensional correlation spectroscopy with ResNet-34 for SSC and firmness classification, followed by integrated quality grading. Visible/near-infrared (Vis/NIR) diffuse reflectance spectra of 600 Xinmei plum samples were collected over 400–1000nm, and reference SSC and firmness values were measured. One-dimensional spectra were transformed into two-dimensional and three-dimensional correlation spectra using single-quality-gradient perturbation, and synchronous, asynchronous, and integrated spectral representations were compared. Using ResNet-34 as the principal backbone, the 3D-COS synchronous representations achieved validation accuracies of 95.33% for SSC and 91.95% for firmness, respectively, outperforming the corresponding 1D-CNN baselines. Feature visualization using t-SNE and Grad-CAM showed improved extraction and separation of discriminative spectral information related to quality levels. The principal practical innovation was a limiting-factor-based fusion strategy that converted the separate SSC and firmness predictions into a unified commercial grade, achieving an integrated grading accuracy of 95.00% and reducing severe cross-grade misclassification. These results indicate that multidimensional correlation spectroscopy coupled with deep residual learning is effective for nondestructive evaluation and grading of Xinmei plum quality.
Grape juice is widely consumed in Brazil and worldwide, but authenticity control and the detection of adulteration remain challenging. This work examined the metabolomic composition of grape juices from southern Brazil and evaluated the use of 1H Nuclear Magnetic Resonance (NMR)-based metabolomics for detecting adulteration with apple juice. Sixty-five samples were analyzed, including authentic grape juices, apple juices, and commercial juice formulations. 1H NMR spectra were acquired on a 300MHz NMR spectrometer and then processed for spectral and chemometric analyses. Principal component analysis (PCA), permutational multivariate analysis of variance (PERMANOVA), and partial least squares discriminant analysis (PLS-DA) supported the discrimination between grape and apple juices. Malic acid at 2.59 ppm was the main discriminant metabolite associated with apple juice, whereas carbohydrate-related signals were primarily linked to formulation and processing effects. Significant differences among juice categories were verified by ANOVA followed by Tukey’s post hoc test (p < 0.05). The results indicate that 1H NMR-based metabolomics combined with chemometric analysis may provide a fast, reproducible, and non-destructive method for grape juice authentication and adulteration detection, with possible application in routine quality control and food authenticity verification.
Vanillin is widely used in food and industrial applications, but excessive intake of vanillin can pose health risks. Therefore, the development of novel detection strategies with high sensitivity and selectivity is essential for the accurate identification of vanillin. Ethyl (Z)-2-cyano-4-oxo-3,4-diphenylbut-2-enoate (ECOD), a novel fluorescent probe, was synthesized by an elimination reaction using ethyl cyanoacetate, benzil, and aqueous ammonia as raw materials. The probe exhibited maximum fluorescence excitation and emission wavelengths at 300 and 340nm, respectively, with a quantum yield of 22.6%. ECOD was found to have high selectivity and sensitivity for the detection of vanillin. This was attributed to ECOD binding to vanillin, which induced conformational changes and triggered intermolecular charge transfer (ICT), ultimately resulting in fluorescence quenching. Therefore, a fluorescence-based method was established for the quantification of vanillin in biscuits, candy, and milk sauce. The linear detection range was 1.0 to 100µM and the detection limit was 0.33µM. The recoveries ranged from 93.2% to 113%, with an RSD ranging from 2.8% to 10%. These results indicated that ECOD had significant application potential in food analysis, suggesting that the fluorescence method based on ECOD could provide a novel approach for rapid vanillin determination in complex food matrices.
Marine microalgae are highly relevant as a likely source of new innovative foods with interesting nutritional and functional characteristics. In this study, several microalgae from the ALGARED collection, extracted from the Andalusian Atlantic shoreline, were screened for their protein content and the antioxidant potential of their protein concentrates and hydrolyzates. Additionally, the in vitro digestibility and antitumor activity of selected species were evaluated. The highest extraction yields and protein content were found in the phylum Cyanophyta (Cyanobium and Synechococcus), while the highest antioxidant activity was observed in species from the phylum Chlorophyta (Tetraselmis, Picochlorum, Chlamydomonas). Other phyla also showed promising results, including Nannochloropsis and Cylindrotheca from Heterokontophyta, Rhodomonas from Cryptophyta, and Prymnesium from Haptophyta. The in vitro digestibility of selected Chlorophyta species (Tetraselmis, Picochlorum) supported their preliminary potential value as sources of bioaccessible antioxidant compounds. All analyzed protein concentrates exhibited moderate antiproliferative activity (IC50 ranging from 121.7 to 180.8µg/mL). In conclusion, the chemical screening of various microalgae from the ALGARED collection has led to the characterization of several species with promising nutritional and biofunctional properties. These microalgae could be further explored for the elaboration of high-protein functional ingredients and novel foodstuffs.
Food fraud due to water adulteration in orange juice is a significant concern for the beverage industry. This study aimed to develop a rapid and non-invasive computer-vision method based on deep learning to detect and classify water-adulteration levels ranging from 1% to 15% in three orange juice products representing freshly squeezed juice, juice from concentrate, and orange nectar. High-resolution images were acquired under two shutter-speed-based exposure conditions: the higher-exposure acquisition condition (1/30s) and the lower-exposure acquisition condition (1/250s), while maintaining constant scene illumination, and were analyzed using ResNet50 convolutional neural networks. In the image-level hold-out test, the model trained with images acquired at 1/250s achieved an accuracy of 88.3%, whereas the model trained at 1/30s reached 83.1%. To provide a more stringent assessment of sample-level transferability, the previously trained and fixed 1/250s model was subsequently evaluated using 240 independently prepared and blindly coded samples obtained from subsequent purchases of the same commercial products. This independent blind validation achieved a 24-class accuracy of 86.7%. Misclassifications occurred predominantly between adjacent or closely related water-adulteration levels within the same juice product. When the independent-validation predictions were collapsed into a binary pure-versus-adulterated screening task, the model achieved 100.0% sensitivity, 93.3% specificity, 99.2% accuracy, and 96.7% balanced accuracy. These findings support the feasibility of the method as a rapid pre-screening tool for detecting visible-image patterns associated with controlled water dilution and show that detecting the presence of adulteration is more reliable than assigning an exact value to closely spaced adulteration levels. The independent validation provides evidence of transferability to newly prepared samples from subsequent purchases of the same products. However, further validation across additional brands, production batches, orange origins, seasons, and acquisition environments is required before broader applicability can be established.
Ginsenosides are important quality markers of Panax ginseng, but their simultaneous quantification remains difficult because many structurally similar compounds are not sufficiently resolved chromatographically. Although LC–MS methods offer broad analyte coverage, their cost and operational complexity limit routine application. Conventional LC methods using DAD or ELSD are more accessible, but their simultaneous quantification capacity has generally remained limited to approximately 22 ginsenosides. Our previous C18-based LC-DAD-ELSD method expanded this coverage to 41 ginsenosides, but insufficient separation among several analytes remained a limitation. In this study, a C30-based LC-DAD-ELSD method was developed to improve chromatographic separation while preserving the broad analyte coverage of the previous platform. Compared with the C18-based method, the C30 column increased the number of compounds with resolution ≥ 1.5 from 22/41 to 40/41 and reduced peak overlap. Correspondingly, 27 ginsenosides showed reduced LoD and/or LoQ values, with up to five-fold lower LoQ values for selected compounds. Application to cultivated and wild-simulated ginseng samples enabled reliable quantification of 20 ginsenosides and revealed distinct compositional differences among sample types. Collectively, the results support the use of the C30-based LC-DAD-ELSD method as an improved non-MS approach for broad-panel ginsenoside analysis in ginseng matrices.
Oviductus ranae is rich in protein and fat, serving as a significant functional food. This study employed label-free DIA rapid quantitative proteomics technology to examine the differences in protein composition of Oviductus Ranae processed by negative pressure airflow combined desiccator drying method(NPAD） in comparison to traditional drying methods(natural drying and hot-air drying), as well as its effects on physicochemical properties, microstructure, and protein structure. The findings indicate that, in comparison to traditional drying techniques, NPAD minimizes the loss of nutritional quality while better preserving the appearance and internal structure. The α-helix content in the secondary structure was 39.62% for NPAD, 38.71% for ND, and only 25.14% for HAD. Furthermore, proteomics techniques were employed to elucidate the molecular mechanisms by which key differential proteins (DEPs) influence trait quality.Notable DEPs, including glutathione S-transferase ω, β-enolase, fructose-diphosphate aldolase, paralbumin, actin β chain.These proteins sensitively reflect the variations induced by the current drying process and can be regarded as candidate indicator proteins for responding to different drying conditions. These results provide mechanistic insights into the quality changes of Oviductus ranae during the drying process and establish theoretical and practical foundations for selecting optimal drying methods to produce high-quality Oviductus ranae.
The quality evaluation of Coptidis Rhizoma (CR) is analytically challenging because of its multiple botanical origins and processed forms. This study developed a chemometric analytical workflow for botanical origin discrimination and processing-related taste profiling of CR by integrating infrared spectroscopy, UPLC-Q-TOF-MS/MS fingerprinting, and electronic tongue (E-tongue) analysis. For origin discrimination, FT-IR and NIR spectroscopy were systematically compared using different preprocessing, variable-selection, and supervised classification strategies. NIR models based on second-derivative preprocessing combined with VIP or CARS variable selection achieved 100% classification accuracy for discriminating C. chinensis, C. deltoidea, and C. teeta. UPLC-Q-TOF-MS/MS fingerprinting coupled with CP-ANN modeling identified six alkaloids as origin-related markers, with C. chinensis showing relatively higher levels of most quantified alkaloids. For processed products prepared from C. chinensis (WL), E-tongue analysis clearly differentiated raw and processed samples and identified umami (NMS), saltiness (CTS), and general taste (PKS) as key sensory dimensions. Relative sensor-response analysis further showed that processing induced differential changes across taste-related sensor dimensions, with the bitterness-related sensor showing the smallest relative increase among all sensors, while wine-processed CR exhibited the most pronounced overall sensory shift. Overall, this workflow provides a practical and reproducible analytical strategy for herbal materials with botanical-origin and processing-dependent quality variability.
Fungal infection during wheat storage triggers complex physicochemical changes, including macromolecular degradation and alterations in moisture-related states, which may remain difficult to detect during early stages. In this study, we proposed a physics-based terahertz (THz) dielectric spectroscopic methodology to characterize mold-associated changes in wheat grains. By integrating the Landau-Lifshitz-Looyenga (LLL) effective medium theory with a generalized fractional-order Cole-Cole model, we extracted the intrinsic dielectric properties of wheat grains. Our results identify two major severity-associated dielectric responses: the relaxation time (τ) reflects changes in water-binding dynamics associated with grain matrix alterations, while the high-frequency permittivity (ε∞) is associated with compositional and hydration-related changes, with reducing-substance (DNS) measurements providing complementary biochemical information. By coupling these physically interpretable parameters with a Support Vector Machine (SVM), the proposed framework achieved an internal cross-validation accuracy of 94.2% with a macro-F1 score of 0.950 for mold severity grading. This research provides a physically interpretable THz dielectric framework for grain quality assessment, offering a rapid, non-destructive, and reagent-free approach for evaluating internal moisture-related relaxation dynamics and compositional integrity.
With the rapid expansion of the pet food market, the quality, safety, and nutritional balance of pet feed have drawn increasing attention. Traditional analytical methods, although accurate, are often time‑consuming, involve complex procedures, and rely on chemical reagents, making them unsuitable for rapid quality control in modern production. To address these limitations, this study developed a rapid quantitative approach based on the fusion of portable near‑infrared (NIR) and attenuated total reflection mid‑infrared (ATR‑IR) spectroscopy combined with chemometrics for determining key nutritional components, including crude protein, crude fat, moisture, and lysine, in canine feed. Partial least squares regression (PLSR) models were established, and both low‑level and mid‑level data fusion strategies were evaluated. For crude protein and moisture, the NIR model alone provided the best predictive performance; for crude fat and lysine, the mid‑level fusion spectra achieved the highest accuracy. The proposed method is efficient, non‑destructive, and environmentally friendly, offering reliable technical support for rapid quality monitoring in pet feed production.
Aging enhances meat quality; however, the alterations in quality, flavor compounds, and muscle proteins during the wet aging of deer meat have not been comprehensively investigated. The longissimus thoracis muscle of sika deer was collected immediately after slaughter, sectioned, and vacuum-aged at 3°C for up to 24 days. Changes in color tone, water-holding capacity, umami compounds, and myofibrillar proteins of the aged meat were analyzed. Taste compound analysis revealed that 5’-inosinic acid decreased during the early stages of aging, whereas free amino acids increased significantly after day 7 and reached a maximum on day 15, indicating an enhancement in the umami of deer meat. With no microbiological or sensory evaluations, the results rely solely on biochemical quality; the mK-value, a promising aging indicator, revealed that biochemical freshness declined markedly after 15 days. Western blotting revealed troponin T and desmin fragmentation after 7 days, indicating meat softening had occurred and that these proteins are useful markers for monitoring the aging process in deer meat. The optimal aging period for deer meat based on physicochemical indicators was found to be 7–11 days at 3°C under vacuum packaging. These findings provide fundamental data to monitor the aging process to produce high-quality deer meat.
The presence of cartilage in poultry is considered undesirable. Four novel tryptic cartilage marker peptides derived from chicken collagen IIα1 and IXα1 were identified using LC-QToF-MS/MS in a comprehensive identification process and were not detected in tendons and skin. It was shown that the solely qualitative detection of cartilage-derived peptides is not sufficient for the identification of mechanically separated meat (MSM), as these peptides were – albeit in lower concentrations – also detected in meat intended for further processing into meat products (non-MSM). Therefore, a quantitative method (LC-QQQ-MS/MS) based on matrix-matched calibration using chicken fillets spiked with different concentrations of defatted cartilage powder (0.01, 0.05, 0.1, 0.2, or 0.5%) was developed. A satisfactory level of homogeneity of the sample material for routine analysis was achieved by extensively optimizing the sample preparation procedures, including freeze-drying, Soxhlet defatting with petroleum ether, subsequent grinding, and a substantial increase in the sample weight. The multiple reaction monitoring analysis of the cartilage-specific peptide VMQEQLSQLAASLR from collagen IXα1 exhibited the best quantification performance with low relative errors (%RE) below 15% at all cartilage concentration levels. The method developed provides a basis for evaluating the potential discrimination between MSM and non-MSM in commercial chicken raw materials.
Single-nucleotide polymorphisms (SNPs) are distinguished by abundance, stability, and independence from developmental stages and environments, which makes them powerful tools for quality control of Chinese herbal medicines. In this review, we systematically examine the fundamental SNP detection platforms, with a focus on the innovative breakthroughs driven by CRISPR-based technology. These advances primarily involve optimizing components of the CRISPR/SNP system, including engineering high-fidelity Cas proteins, engineering crRNA sequences, introducing synthetic mismatches into the spacer region, and overcoming the protospacer adjacent motif (PAM) sequence limitations. Collectively, these strategies enhance detection specificity and accuracy, paving the way for applications in Chinese herbal medicine. Furthermore, we synthesize the classic applications of SNP detection, propose its prospects in Chinese herbal medicine, and simultaneously explore current challenges and future directions. It aims to provide a framework for advancing precise identification and quality control of Chinese herbal medicine.
Pig castration status may influence fat deposition, tissue composition, and sensory quality, creating a need for rapid verification methods. This study evaluated near-infrared spectroscopy (NIR) combined with machine learning to classify pork samples from castrated and intact male pigs. Spectra were collected from 200 subcutaneous adipose-tissue samples (80 castrated and 120 intact) and processed using four pipelines involving standard normal variate transformation, multiplicative scatter correction, Savitzky–Golay filtering, derivatives, and detrending. Random forest, support vector machine (SVM), and k-nearest neighbors models were compared. Accuracy was the primary metric, with weighted F1-score, cross-validation performance, and AUC used for complementary evaluation. The best combination was standard normal variate transformation, Savitzky–Golay filtering, detrending, and SVM, which achieved 96.00% accuracy, a 96.00% weighted F1-score, and a 97.38% cross-validation score. Repeated stratified nested cross-validation supported SVM performance, and permutation-importance analysis identified six informative wavelength regions. These findings demonstrate the feasibility of NIR-based castration-status classification. However, because samples came from one farm and slaughter batch and no independent external validation or chemical or sensory reference measurements were available, the study should be considered proof of concept. Future work should include multi-source external validation and portable or online implementation.
