Search

Breast milk is the safest and most perfect natural food for infant growth and development. As one of the most important components in breast milk, the milk fat globule membrane (MFGM) is a 3-layer membrane structure surrounding milk fat globules (MFG). This unique structure not only maintains the stability of milk but also plays an important role in the digestive and metabolic processes of infants. In this article, we introduce the reader to the composition and structural specificity of MFGM, review the sequential digestion of MFGM depending on several enzymes in the mouth, stomach and intestine of healthy infants, and elaborate on the interaction mechanism between MFGM and various enzymes, in order to provide a reference for relevant studies.

ObjectiveThis study aims to evaluate the therapeutic efficacy of high-dose corticosteroid therapy in children diagnosed with epileptic encephalopathy with spike-and-wave activation in sleep (EE-SWAS), investigate associated clinical indicators influencing treatment outcomes, and establish a predictive model for recurrence.MethodsChildren diagnosed with EE-SWAS who received high-dose corticosteroid therapy were categorized into responder group and non-responder group. Data on clinical parameters, electroencephalogram (EEG) features, and serum cytokine levels were collected. Six months post-treatment, the effectively treated children were further stratified into recurrence and non-recurrence groups. Risk factors for poor outcomes following corticosteroid therapy were identified using univariate analysis. Multivariate logistic regression analysis was then employed to determine independent factors influencing the recurrence of corticosteroid therapy, which facilitated the development of a predictive model.ResultsThe study included 48 children, with 33 cases in the responder group (effective rate = 68.8%) and 15 cases in the non-responder group. The responder group exhibited an older onset age of electrical status epilepticus in sleep (ESES) and higher proportions of combined benzodiazepines (BZDs) use (P

Phosphoglycerate mutase/protein phosphatase (PGAM5)-mediated cell death plays an important role in multiple liver diseases. However, few studies have confirmed the regulatory mechanism of melatonin acting on PGAM5-mediated cell death in the context of liver ischemia-reperfusion (I/R) injury. The liver I/R injury model and cell hypoxia-reoxygenation model were established after melatonin pretreatment. Liver injury, cell activity, cell apoptosis, oxidative stress index, and PGAM5 protein expression were detected. To investigate the role of PGAM5 in melatonin-mediated liver protection during I/R injury, PGAM5 silencing, and overexpression were performed before melatonin pretreatment. Our results indicated that PGAM5 was significantly elevated by I/R injury, and predominantly localized in the necrosis area. However, treatment with melatonin blocked PGAM5 activation and conferred a survival advantage of hepatocytes in liver I/R injury, similar to the results achieved by silencing PGAM5. In terms of mechanism, we illustrated that activated PGAM5 promoted mitochondrial permeability transition pore (mPTP) opening, and administration of melatonin inhibited mPTP opening and interrupted hepatocytes death via blocking PGAM5. Our data indicated that the PGAM5-mPTP axis is responsible for I/R-induced liver injury. In contrast, melatonin supplementation blocked the PGAM5-mPTP axis and thus decreased cell death, providing a protective advantage to hepatocytes in I/R. These results established a new paradigm in melatonin-mediated hepatocyte protection under the burden of I/R attack. [ABSTRACT FROM AUTHOR]

A comprehensive understanding of the characteristics and formation mechanisms of groundwater in mining areas is essential for the effective prevention of coal mine water and the rational management of groundwater resources. The objective of this study was to examine the hydrogeochemical characteristics and evolution of Ordovician groundwater in the Zhuozishan coal mine, located in the northwest region of China. A total of 34 groundwater samples were collected for hydrogeochemical analyses and the investigation of their evolution processes, with the aid of a piper trilinear diagram, a Gibbs diagram, and an ion ratio diagram. The results indicate that the concentration of sodium (Na+), potassium (K+), bicarbonate ( HCO 3 − ), chloride (Cl−), sulphate ( SO 4 2 − ), total dissolved solids (TDS), and pH increases from the recharge area to the discharge area, whereas the concentration of calcium (Ca2+) and magnesium (Mg2+) decreases. The hydrogeochemical characteristics of the runoff from Zhuozishan to Gongdeer coalfield and further southward display a notable north–south directional change. The groundwater process is primarily controlled by rock weathering action and cation exchange, with Na+ and K+ deriving primarily from cation exchange and only to a minor extent from halite dissolution. In conclusion, the northern part of the coalfield is characterised by a geological structure that creates a retention area with groundwater, resulting in an unordered runoff process with a complex formation mechanism. The middle region is devoid of geological constraints that would alter the flow direction, thus simplifying the process of groundwater formation. In contrast, the southern area experiences an increase in strata depth and fault blocking, which creates a retention zone, thereby rendering the groundwater formation process more complex. This research contributes to the effective management of groundwater resources in this coalfield and other mining sites. [ABSTRACT FROM AUTHOR]

As deep learning has evolved, larger and deeper neural networks are currently a popular trend in both natural language processing tasks and computer vision tasks. With the increasing parameter size and model complexity in deep neural networks, it is also necessary to have more data available for training to avoid overfitting and to achieve better results. These facts demonstrate that training deep neural networks takes more and more time. In this paper, we propose a training acceleration method based on gradually freezing the parameters during the training process. Specifically, by observing the convergence trend during the training of deep neural networks, we freeze part of the parameters so that they are no longer involved in subsequent training and reduce the time cost of training. Furthermore, an adaptive freezing algorithm for the control of freezing speed is proposed in accordance with the information reflected by the gradient of the parameters. Concretely, a larger gradient indicates that the loss function changes more drastically at that position, implying that there is more room for improvement with the parameter involved; a smaller gradient indicates that the loss function changes less and the learning of that part is close to saturation, with less benefit from further training. We use ViTDet as our baseline and conduct experiments on three remote sensing target detection datasets to verify the effectiveness of the method. Our method provides a minimum speedup ratio of 1.38×, while maintaining a maximum accuracy loss of only 2.5%. [ABSTRACT FROM AUTHOR]