Aiming at the problems of low detection accuracy for small-sized defects,weak multi-scale fea-ture extraction ability and low anomaly segmentation accuracy of existing industrial anomaly detection algo-rithms,an industrial anomaly detection network combining high-frequency residual guidance and multi-scale attention feature fusion was proposed.Firstly,aiming at the problem of high-frequency detail loss caused by traditional full-frequency processing,a frequency-domain separation strategy was designed.Gaussian kernel filtering was utilized to extract high-frequency residual features,enhancing the network's detection ability for minor anomalies.Secondly,aiming at the problems of insufficient representation abili-ty of complex textures and low discrimination between anomalies and backgrounds in conventional convolu-tional networks,a globally enhanced multi-scale attention module GEMA is embedded in the encoder stage of the discriminative network.It captures multi-scale local information in the horizontal and vertical directions through parallel dual-path,enhancing the salient features at different spatial positions.Improve the feature discriminability in complex texture backgrounds;Finally,in the decoder stage of the discrimi-nant network,the coordinate attention module CoordAtt is integrated.By decomposing the coordinate ax-es and dynamically modulating the feature weights,precise spatial positioning of abnormal areas is achieved.Experiments show that on the MVTec AD public dataset,the average AUROC at the image level of the improved model is 98.6％,and the average AUROC and AP at the pixel level are 97.6％and 73.2％respectively,effectively improving the effect of industrial anomaly detection.
Many studies have shown that the development of allergic asthma is associated with intestinal microbiota dysbiosis. Based on the gut-lung axis theory, probiotic intervention may be a potential strategy for respiratory diseases. Lacticaseibacillus paracasei K56 was reported to regulate gut microbiota in mice. However, its effect on allergic asthma has not been reported. In this study, we investigated the effect of the K56 on ovalbumin-induced asthma and its possible mechanisms. Our results showed that K56 reduced asthma symptoms and inflammatory cell infiltration in the lung of asthmatic mice. And K56 regulated the differentiation of helper T cells in the lung and intestine. Results from 16S rRNA gene sequencing showed that K56 prophylaxis significantly elevated the richness of Akkermansia and Burkholderia. Meanwhile, K56 abrogated the decrease in the levels of short-chain fatty acids (SCFAs) in the feces. In addition, a rebound in the mRNA expression of the SCFA receptors was observed after K56 prophylaxis. Notably, the regulatory T cell (Treg) frequencies and the FFAR3 levels were positively correlated. These results suggest that K56 could attenuate asthma, possibly by modulating the intestinal microbiota and regulating Treg differentiation through SCFA metabolized. Our study showed that K56 may be used as a probiotic to prevent pulmonary inflammation and dysbiosis of the intestinal microbiota in allergic asthma.
Objective To establish a method for determining reasonable control levels of perchlorate in raw materials for infant formula by applying a probabilistic model.Methods Based on the levels of perchlorate in the raw materials of infant formula and the constituent ratio of ingredients,a probabilistic model between raw materials and final products was established using Monte Carlo simulation and the key influencing raw materials were identified through sensitivity analysis.The content threshold of perchlorate in the key raw material,which is the reasonable control level,was found by using step-back method according to the limit of perchlorate in infant formula.Results At the initial levels of perchlorate contamination in raw milk,milk powder,lactose,whey protein powder and other ingredients collected in this study,the probability that the perchlorate level in the final product exceeded the limit value set by the European Union,which is 1.6%when a certain formula is used to manufacture infant formula.Sensitivity analysis showed that raw milk was the ingredient that has the greatest impact on the perchlorate content in the final product.Given that the perchlorate in other raw materials remains the current level,the perchlorate content in raw milk should be controlled below 43.6 µg/kg,to ensure a probability exceeding 99.9%that the perchlorate content in infant formula powder remains below the regulatory limit value.When taking this value as the risk control level for perchlorate in raw milk,the current ratio of raw milk samples exceeding this level was approximately 3.5%.Conclusion The method established in this study can effectively simulate the probability distributions of stable chemical contaminants when transferring from raw materials to the final product and identify the most influential ingredient.By applying this method,the reasonable control level of perchlorate in raw materials of infant formula can be proposed,providing a scientific basis for infant formula enterprises to prevent and control risks associated with raw materials.
A method for the detection of 1,2-propanediol in milk by gas chromatography-mass spectrometry(GC-MS)was established by optimizing pretreatment and chromatographic conditions.The results showed that the samples were extracted with acetonitrile,purified by dispersion solid-phase extraction,separated by DB-WAX chromatographic column(30 m×0.25 mm×0.25 μm)with a split ratio of 10∶1,determined by GC-MS using the se-lective ion monitoring(SIM)mode,and quantified by the external standard method.Under these conditions,1,2-propanediol had a good linear rela-tionship within the mass concentration range of 1.0-50 μg/ml,with a correlation coefficient(R2)＞0.999.The detection limit(LOD)of 1,2-propanediol in milk was 0.005 g/kg,and the limit of quantification(LOQ)was 0.010 g/kg.The recovery rate of standard addition was 93.67％-100.67％,and the relative standard deviation(RSD)of the precision test results was 1.23％-3.91％.This method featured simple pretreatment,strong resistance to matrix interference,accurate results,and minimal contamination and damage to the instrument.It had certain promotion value and could be used for the de-tection of 1,2-propanediol in milk.
Objective:To conduct comparative analysis of the dynamic trend of breast milk lipid composition across different lactation stages in mothers of preterm and term infants with the aim of providing a basis for optimizing feeding strategies for preterm neonates.Methods:From January 2023 to May 2024,nine preterm mother-infant dyads and ten term mother-infant dyads were recruited from Northwest Women's and Children's Hospital and two tertiary hospitals in Gansu province.Breast milk samples were collected at three time points:days 0-7 postpartum(colostrum),days 8-14(transitional milk),and 2 months(mature milk).High-performance liquid chromatography with evaporative light scattering detection(HPLC-ELSD),liquid chromatography-tandem quadrupole mass spectrometry(LC-qMS)and gas chromatography with flame ionization detection(GC-FID)were used to quantify total fat,structural lipids,gangliosides,phospholipids,and fatty acids in breast milk.Results:The lipid composition and its dynamic changes in breast milk for preterm infants significantly differed from those in breast milk for term infants.In preterm breast milk,total phospholipids,phosphatidylcholine(PC),and phosphatidylserine(PS)levels remained stable throughout the lactation period,whereas those in term breast milk gradually decreased.Total phospholipids,PC and sphingomyelin(SM)levels in preterm colostrum,as well as SM levels in preterm transitional and mature milk were lower than their respective counterparts in term breast milk.Conversely,phosphatidylethanolamine(PE)content was higher in preterm mature milk than in its term counterpart.Fatty acid analysis revealed that eicosatrienoic acid(20:3n3)concentrations in preterm colostrum and transitional milk,as well as eicosapentaenoic acid(EPA,20:5n3)levels in preterm transitional and mature milk were significantly lower than their counterparts in term breast milk.However,no statistically significant differences were observed between preterm and term breast milk in docosahexaenoic acid(DHA,22:6n3)or arachidonic acid(ARA,20:4n6)concentrations across lactation stages.Conclusion:The lipid composition of human milk for preterm infants remains relatively stable:the levels of total phospholipids,PC,and PS remain stable,likely in response to the nutritional requirements of preterm infants,while PE levels in mature milk markedly increase,which is maybe related to the neurodevelopmental requirements.Additionally,the levels of key long-chain unsaturated fatty acids(e.g.,EPA)are lower in preterm than in term breast milk,underscoring the need for targeted nutritional interventions to address the specific demands during the rapid developmental phases.
