Soybean isolate protein complexes with different concentrations of inulin by ultrasound treatment: Structural and functional properties.

The effects of ultrasound and different inulin (INU) concentrations (0, 10, 20, 30, and 40 mg/mL) on the structural and functional properties of soybean isolate protein (SPI)-INU complexes were hereby investigated. Fourier transform infrared spectroscopy showed that SPI was bound to INU via hydrogen bonding. All samples showed a decreasing and then increasing trend of α-helix content with increasing INU concentration. SPI-INU complexes by ultrasound with an INU concentration of 20 mg/mL (U-2) had the lowest content of α-helix, the highest content of random coils and the greatest flexibility, indicating the proteins were most tightly bound to INU in U-2. Both UV spectroscopy and intrinsic fluorescence spectroscopy indicated that it was hydrophobic interactions between INU and SPI. The addition of INU prevented the exposure of tryptophan and tyrosine residues to form a more compact tertiary structure compared to SPI alone, and ultrasound caused further unfolding of the structure of SPI. This indicated that the combined effect of ultrasound and INU concentration significantly altered the tertiary structure of SPI. SDS-PAGE and Native-PAGE displayed the formation of complexes through non-covalent interactions between SPI and INU. The ζ-potential and particle size of U-2 were minimized to as low as -34.94 mV and 110 nm, respectively. Additionally, the flexibility, free sulfhydryl groups, solubility, emulsifying and foaming properties of the samples were improved, with the best results for U-2, respectively 0.25, 3.51 µmoL/g, 55.51 %, 269.91 %, 25.90 %, 137.66 % and 136.33 %. Overall, this work provides a theoretical basis for improving the functional properties of plant proteins.

Ecotoxicological impact of naproxen on Eisenia fetida: Unraveling soil contamination risks and the modulating role of microplastics.

Soils represent crucial sinks for pharmaceuticals and microplastics, making them hotspots for pharmaceuticals and plastic pollution. Despite extensive research on the toxicity of pharmaceuticals and microplastics individually, there is limited understanding of their combined effects on soil biota. This study focused on the earthworm Eisenia fetida as test organism to evaluate the biotoxicity and bioaccumulation of the typical pharmaceutical naproxen and microplastics in earthworms. Results demonstrated that high concentrations of naproxen (100 mg kg-1) significantly increased the malondialdehyde (MDA) content, inducing lipid peroxidation. Even though the low exposure of naproxen exhibits no significant influence to Eisenia fetida, the lipid peroxidation caused by higher concentration than environmental relevant concentrations necessitate attention due to temporal and spatial concentration variability found in the soil environment. Meanwhile, microplastics caused oxidative damage to antioxidant enzymes by reducing the superoxide dismutase (SOD) activity and MDA content in earthworms. Metabolome analysis revealed increased lipid metabolism in naproxen-treated group and reduced lipid metabolism in the microplastic-treated group. The co-exposure of naproxen and microplastics exhibited a similar changing trend to the microplastics-treated group, emphasizing the significant influence of microplastics. The detection of numerous including lipids like 17-Hydroxyandrostane-3-glucuronide, lubiprostone, morroniside, and phosphorylcholine, serves to identify potential biomarkers for naproxen and microplastics exposure. Additionally, microplastics increased the concentration of naproxen in earthworms at sub-organ and subcellular level. This study contributes valuable insights into the biotoxicity and distribution of naproxen and microplastics in earthworms, enhancing our understanding of their combined ecological risk to soil biota.

T2 Relaxation Time in Extraocular Muscles of Patients with Mild Thyroid- Associated Ophthalmopathy: Comparing T2 Mapping With and Without Fat Suppression Using Different Measurement Methods.

OBJECTIVE: This study aimed to compare the parametric value of T2 with and without fat suppression (FS) on T2 mapping for the evaluation of extraocular muscles (EOMs) in mild thyroid-associated ophthalmopathy (TAO). METHODS: We prospectively recruited 44 consecutive patients with mild TAO seen between May 2020 and October 2022 and 26 healthy controls with no history of eye- or thyroid-related or other autoimmune diseases. Patients with mild TAO were subdivided into active and inactive groups based on their clinical activity scores. The T2 of each EOM was measured over a large and small area of interest on T2-mapping images with and without FS. Receiver operating characteristic (ROC) curves were used to evaluate the diagnostic efficacy of T2 for detecting TAO activity. RESULTS: The T2 was significantly and heterogeneously higher in the active group than in the inactive and control groups (P < 0.05). FS-T2-mapping images had better signal display within and at the edges of the EOMs than those without FS. It was possible to observe high-signal aggregation visible in the periphery of some EOMs, and the central part showed relatively low signals. The maximum T2 measured in small or large areas with and without FS had good diagnostic efficacy for TAO activity, with that of no-FS being better (the area under the ROC curve of the maximum T2 measured in a small area and a large area without FS was 1.0 and 1.0 and P values of < 0.001 and < 0.001, respectively). CONCLUSION: Maximal T2 with or without FS can facilitate the early clinical detection of mild TAO activity. The maximum T2 in a small area can facilitate active staging of patients with mild TAO.

Efficient electrosynthesis of HO2- from air for sulfide control in sewers.

Electrochemically in-situ generation of oxygen and caustic soda is promising for sulfide management while suffers from scaling, poor inactivating capacity, hydrogen release and ammonia escape. In this study, the four-compartment electrochemical cell efficiently captured oxygen molecules from the air chamber to produce HO2- without generating toxic by-products. Meanwhile, the catalyst layer surface of PTFE/CB-GDE maintained a relatively balanced gas-liquid micro-environment, enabling the formation of enduring solid-liquid-gas interfaces for efficient HO2- electrosynthesis. A dramatic increase in HO2- generation rate from 453.3 mg L-1 h-1 to 575.4 mg L-1 h-1 was attained by advancement in operation parameters design (flow channels, electrolyte types, flow rates and circulation types). Stability testing resulted in the HO2- generation rate over 15 g L-1 and the current efficiency (CE) exceeding 85%, indicating a robust stable operational capacity. Furthermore, after 120 mg L-1 HO2- treatment, an increase of 11.1% in necrotic and apoptotic cells in the sewer biofilm was observed, higher than that achieved with the addition of NaOH, H2O2 method. The in-situ electrosynthesis strategy for HO2- represents a significance toward the practical implementation of sulfide abatement in sewers, holding the potential to treat various sulfide-containing wastewater.

AutoNet-Generated Deep Layer-Wise Convex Networks for ECG Classification.

The design of neural networks typically involves trial-and-error, a time-consuming process for obtaining an optimal architecture, even for experienced researchers. Additionally, it is widely accepted that loss functions of deep neural networks are generally non-convex with respect to the parameters to be optimised. We propose the Layer-wise Convex Theorem to ensure that the loss is convex with respect to the parameters of a given layer, achieved by constraining each layer to be an overdetermined system of non-linear equations. Based on this theorem, we developed an end-to-end algorithm (the AutoNet) to automatically generate layer-wise convex networks (LCNs) for any given training set. We then demonstrate the performance of the AutoNet-generated LCNs (AutoNet-LCNs) compared to state-of-the-art models on three electrocardiogram (ECG) classification benchmark datasets, with further validation on two non-ECG benchmark datasets for more general tasks. The AutoNet-LCN was able to find networks customised for each dataset without manual fine-tuning under 2 GPU-hours, and the resulting networks outperformed the state-of-the-art models with fewer than 5% parameters on all the above five benchmark datasets. The efficiency and robustness of the AutoNet-LCN markedly reduce model discovery costs and enable efficient training of deep learning models in resource-constrained settings.

Inducing spin polarization via Co doping in the BiVO4 cell to enhance the built-in electric field for promotion of photocatalytic CO2 reduction.

The efficiency of CO2 photocatalytic reduction is severely limited by inefficient separation and sluggish transfer. In this study, spin polarization was induced and built-in electric field was strengthened via Co doping in the BiVO4 cell to boost photocatalytic CO2 reduction. Results showed that owing to the generation of spin-polarized electrons upon Co doping, carrier separation and photocurrent production of the Co-doped BiVO4 were enhanced. CO production during CO2 photocatalytic reduction from the Co-BiVO4 was 61.6 times of the BiVO4. Notably, application of an external magnetic field (100 mT) further boosted photocatalytic CO2 reduction from the Co-BiVO4, with 68.25 folds improvement of CO production compared to pristine BiVO4. The existence of a built-in electric field (IEF) was demonstrated through density functional theory (DFT) simulations and kelvin probe force microscopy (KPFM). Mechanism insights could be elucidated as follows: doping of magnetic Co into the BiVO4 resulted in increased the number of spin-polarized photo-excited carriers, and application of a magnetic field led to an augmentation of intrinsic electric field due to a dipole shift, thereby extending carrier lifetime and suppressing charges recombination. Additionally, HCOO- was a crucial intermediate in the process of CO2RR, and possible pathways for CO2 reduction were proposed. This study highlights the significance of built-in electric fields and the important role of spin polarization for promotion of photocatalytic CO2 reduction.

l-norleucine on high glucose-induced insulin sensitivity and mitochondrial function in skeletal muscle cells.

l-norleucine, an isomer of leucine, stimulates the anabolic process of insulin. However, it is not known if and how it improves insulin sensitivity and insulin resistance. This experiment describes the generation of an insulin resistance model using high glucose-induced cells and the administration of 1.0 mmol/L l-norleucine for 48 h, to observe the effects on metabolism and gene expression in skeletal muscle cells. The results showed that l-norleucine significantly increased mitochondrial ATP content, decreased the amount of reactive oxygen species (ROS) and promoted the expression of mitochondrial generation-related genes TFAM, AMPK, PGC-1α in cells under high glucose treatment; at the same time, l-norleucine also increased glucose uptake, suggesting that l-norleucine increased insulin sensitivity and improved insulin resistance. This study suggesting that l-norleucine improves insulin resistance by ameliorating oxidative stress damage of mitochondria, improving mitochondrial function, and improving insulin sensitivity in skeletal muscle cell caused by high glucose, rather than by altering mitochondrial efficiency.

Differences in inpatient performance of public general hospitals following implementation of a points-counting payment based on diagnosis-related group: a robust multiple interrupted time series study in Wenzhou, China.

OBJECTIVES: This study measures the differences in inpatient performance after a points-counting payment policy based on diagnosis-related group (DRG) was implemented. The point value is dynamic; its change depends on the annual DRGs' cost settlements and points of the current year, which are calculated at the beginning of the following year. DESIGN: A longitudinal study using a robust multiple interrupted time series model to evaluate service performance following policy implementation. SETTING: Twenty-two public general hospitals (8 tertiary institutions and 14 secondary institutions) in Wenzhou, China. INTERVENTION: The intervention was implemented in January 2020. OUTCOME MEASURES: The indicators were case mix index (CMI), cost per hospitalisation (CPH), average length of stay (ALOS), cost efficiency index (CEI) and time efficiency index (TEI). The study employed the means of these indicators. RESULTS: The impact of COVID-19, which reached Zhejiang Province at the end of January 2020, was temporary given rapid containment following strict control measures. After the intervention, except for the ALOS mean, the change-points for the other outcomes (p<0.05) in tertiary and secondary institutions were inconsistent. The CMI mean turned to uptrend in tertiary (p<0.01) and secondary (p<0.0001) institutions compared with before. Although the slope of the CPH mean did not change (p>0.05), the uptrend of the CEI mean in tertiary institutions alleviated (p<0.05) and further increased (p<0.05) in secondary institutions. The slopes of the ALOS and TEI mean in secondary institutions changed (p<0.05), but not in tertiary institutions (p>0.05). CONCLUSIONS: This study showed a positive effect of the DRG policy in Wenzhou, even during COVID-19. The policy can motivate public general hospitals to improve their comprehensive capacity and mitigate discrepancies in treatment expenses efficiency for similar diseases. Policymakers are interested in whether the reform successfully motivates hospitals to strengthen their internal impetus and improve their performance, and this is supported by this study.

Sulfur-containing substances in sewers: Transformation, transportation, and remediation.

Sulfur-containing substances in sewers frequently incur unpleasant odors, corrosion-related economic loss, and potential human health concerns. These observations are principally attributed to microbial reactions, particularly the involvement of sulfate-reducing bacteria (SRB) in sulfur reduction process. As a multivalent element, sulfur engages in complex bioreactions in both aerobic and anaerobic environments. Organic sulfides are also present in sewage, and these compounds possess the potential to undergo transformation and volatilization. In this paper, a comprehensive review was conducted on the present status regarding sulfur transformation, transportation, and remediation in sewers, including both inorganic and organic sulfur components. The review extensively addressed reactions occurring in the liquid and gas phase, as well as examined detection methods for various types of sulfur compounds and factors affecting sulfur transformation. Current remediation measures based on corresponding mechanisms were presented. Additionally, the impacts of measures implemented in sewers on the subsequent wastewater treatment plants were also discussed, aiming to attain better management of the entire wastewater system. Finally, challenges and prospects related to the issue of sulfur-containing substances in sewers were proposed to facilitate improved management and development of the urban water system.

Association of different cell types and inflammation in early acne vulgaris.

Acne vulgaris, one of the most common skin diseases, is a chronic cutaneous inflammation of the upper pilosebaceous unit (PSU) with complex pathogenesis. Inflammation plays a central role in the pathogenesis of acne vulgaris. During the inflammatory process, the innate and adaptive immune systems are coordinately activated to induce immune responses. Understanding the infiltration and cytokine secretion of differential cells in acne lesions, especially in the early stages of inflammation, will provide an insight into the pathogenesis of acne. The purpose of this review is to synthesize the association of different cell types with inflammation in early acne vulgaris and provide a comprehensive understanding of skin inflammation and immune responses.

Feasibility of wearable monitors to detect heart rate variability in children with hand, foot and mouth disease.

Hand foot and mouth disease (HFMD) is caused by a variety of enteroviruses, and occurs in large outbreaks in which a small proportion of children deteriorate rapidly with cardiopulmonary failure. Determining which children are likely to deteriorate is difficult and health systems may become overloaded during outbreaks as many children require hospitalization for monitoring. Heart rate variability (HRV) may help distinguish those with more severe diseases but requires simple scalable methods to collect ECG data.We carried out a prospective observational study to examine the feasibility of using wearable devices to measure HRV in 142 children admitted with HFMD at a children's hospital in Vietnam. ECG data were collected in all children. HRV indices calculated were lower in those with enterovirus A71 associated HFMD compared to those with other viral pathogens.HRV analysis collected from wearable devices is feasible in a low and middle income country (LMIC) and may help classify disease severity in HFMD.

Biomass-Derived Carbon Quantum Dots Chemical Bonding with Sn-Doped Bi2 O2 CO3 Endowed Fast Carriers' Dynamics and Stabilized Oxygen Vacancies for Efficient Decontamination of Combined Pollutants.

Surface defects on photocatalysts could promote carrier separation and generate unsaturated sites for chemisorption and reactant activation. Nevertheless, the inactivation of oxygen vacancies (OVs) would deteriorate catalytic activity and limit the durability of defective materials. Herein, bagasse-derived carbon quantum dots (CQDs) are loaded on the Sn-doped Bi2 O2 CO3 (BOC) via hydrothermal procedure to create Bi─O─C chemical bonding at the interface, which not only provides efficient atomic-level interfacial electron channels for accelerating carriers transfer, but also enhances durability. The optimized Sn-BOC/CQDs-2 achieves the highest photocatalytic removal efficiencies for levofloxacin (LEV) (88.7%) and Cr (VI) (99.3%). The elimination efficiency for LEV and Cr (VI) from the Sn-BOC/CQDs-2 is maintained at 55.1% and 77.0% while the Sn-BOC is completely deactivated after four cycle tests. Furthermore, the key role of CQDs in stabilization of OVs is to replace OVs as the active center of H2 O and O2 adsorption and activation, thereby preventing reactant molecules from occupying OVs. Based on theoretical calculations of the Fukui index and intermediates identification, three possible degradation pathways of LEV are inferred. This work provides new insight into improving the stability of defective photocatalysts.

Performance evaluation of air-source heat pump based on a pressure drop embedded model.

An air-source heat pump simulation model, accounting for evaporator and condenser pressure drop, has been developed. The model is capable of computing the heat pump's coefficient of performance (COP) under different ambient temperatures and relative humidities above frosting conditions. This research extends an existing iterative simulation method that relies on the equalization of logarithmic mean temperature differences (LMTDs) calculated through two different approaches by adding a pressure drop simulation. Frictional and acceleration pressure drop is considered, computed iteratively. Simulation results for three different refrigerants, R410A, R32 and R290, are compared. The model's accuracy is validated by comparing simulated COP values with measured COP values from the reference heat pump datasheet. The model closely replicates the measured COP values above frosting conditions, with only a slight underestimation of approximately 1.5%. Results show a substantial impact of ambient temperature on the COP. For instance, an ambient temperature of 20 ◦C, compared to 7 ◦C, results in a COP increase of up to 35%, while an ambient temperature of -10 ◦C leads to a 26% reduction in COP. Relative humidity enhances the COP if air moisture condensation becomes possible. Higher condenser capacities negatively affect the COP. The study highlights the differences in pressure drop characteristics between the condenser and the evaporator for the modeled heat pump, with maximum pressure drops of 220 kPa and 50 kPa for the condenser and evaporator, respectively. Additionally, the choice of refrigerant significantly influences pressure drop, with R32 displaying the lowest pressure drop, R410A showing the highest condenser pressure drop, and R290 causing the highest evaporator pressure drop.

Trends in the prevalence of social isolation among middle and older adults in China from 2011 to 2018: the China Health and Retirement Longitudinal Study.

BACKGROUND: Epidemiological studies have shown that social isolation, which is prevalent in older adults, is associated with a range of adverse health outcomes, but the prevalence of and trends in regard to social isolation remain ambiguous in China. The aim of this study was to elucidate the trends regarding the prevalence of social isolation among middle-aged and older adults in China from 2011 to 2018 and to further identify associated risk factors. METHODS: A repeated cross-sectional study, The data were derived from panel sample data of four waves conducted from May 2011 to August 2018 in the nationally representative China Health and Retirement Longitudinal Study (CHARLS) using multistage probability sampling. Social isolation was ascertained by the five item Steptoe Social Isolation Index. The potential covariates were demographic characteristics, lifestyle factors, and health status. Linear-by-linear association was used to assess the trends in regard to social isolation over time under the influence of the potential covariates. Linear-by-linear association and an age-period-cohort analysis were used to explore the trends, and two-level (time, individual) generalized estimating equation models (GEE) linked multivariate binary logistic regression were performed to identify risk factors. RESULTS: A high prevalence of social isolation and a moderate upward trend from 2013 to 2018 were observed among a U-shaped trend prevalence of social isolation from 2011 to 2018 across China, with rates of 38.09% (95% CI = 36.73-39.45) in 2011, 33.66% (32.32-35.00) in 2013, 39.13% (37.59-40.67) in 2015, and 39.95% (38.59-41.31) in 2018 (p < 0.001). The prevalence of social isolation increased with age and educational attainment. Females had a higher prevalence than males. The prevalence of social isolation was found to be significantly lower in pensioners than in non-pensioners between 2011 and 2018 (p < 0.001). The prevalence of social isolation was 38.9%, 34.9%, 38.5%, and 44.08% about three times higher among those who doid not use the Internet and 13.44%, 11.64%, 12.93%, and 16.73% than among those who doid in 2011, 2013, 2015 and 2018 respectively. The participants with short (0-5 h) and long sleep (9 or more hours), and poor self-rated health had a higher prevalence of social isolation than the others. Older age, lower educational attainment, living in a rural region, lack of medical insurance or pension, lack of internet use and poor health were risk factors (p < 0.05). CONCLUSIONS: We found a U-shaped prevalence of social isolation trends from 2011 to 2018 and revealed increasing trends from 2013 to 2018 among middle-aged and older adults in China. The findings of the study highlight the urgent need for interventions to reduce social isolation including improving sleep quality and internet skills. Disadvantaged groups in terms of age, economic status, and health status should be the focus of such interventions, especially in the era of COVID-19.

Data encoding for healthcare data democratization and information leakage prevention.

The lack of data democratization and information leakage from trained models hinder the development and acceptance of robust deep learning-based healthcare solutions. This paper argues that irreversible data encoding can provide an effective solution to achieve data democratization without violating the privacy constraints imposed on healthcare data and clinical models. An ideal encoding framework transforms the data into a new space where it is imperceptible to a manual or computational inspection. However, encoded data should preserve the semantics of the original data such that deep learning models can be trained effectively. This paper hypothesizes the characteristics of the desired encoding framework and then exploits random projections and random quantum encoding to realize this framework for dense and longitudinal or time-series data. Experimental evaluation highlights that models trained on encoded time-series data effectively uphold the information bottleneck principle and hence, exhibit lesser information leakage from trained models.

MiR-431 promotes cardiomyocyte proliferation by targeting FBXO32 expression.

BACKGROUND: The induction of cardiomyocyte (CM) proliferation is a promising approach for cardiac regeneration following myocardial injury. MicroRNAs (miRNAs) have been reported to regulate CM proliferation. In particular, miR-431 expression decreases during cardiac development, according to Gene Expression Omnibus (GEO) microarray data. However, whether miR-431 regulates CM proliferation has not been thoroughly investigated. METHODS: We used integrated bioinformatics analysis of GEO datasets to identify the most significantly differentially expressed miRNAs. Real-time quantitative PCR and fluorescence in situ hybridization were performed to determine the miRNA expression patterns in hearts. Gain- and loss-of-function assays were conducted to detect the role of miRNA in CM proliferation. Additionally, we detected whether miR-431 affected CM proliferation in a myocardial infarction model. The TargetScan, miRDB and miRWalk online databases were used to predict the potential target genes of miRNAs. Luciferase reporter assays were used to study miRNA interactions with the targeting mRNA. RESULTS: First, we found a significant reduction in miR-431 levels during cardiac development. Then, by overexpression and inhibition of miR-431, we demonstrated that miR-431 promotes CM proliferation in vitro and in vivo, as determined by immunofluorescence assays of 5-ethynyl-2'-deoxyuridine (EdU), pH3, Aurora B and CM count, whereas miR-431 inhibition suppresses CM proliferation. Then, we found that miR-431 improved cardiac function post-myocardial infarction. In addition, we identified FBXO32 as a direct target gene of miR-431, with FBXO32 mRNA and protein expression being suppressed by miR-431. FBXO32 inhibited CM proliferation. Overexpression of FBXO32 blocks the enhanced effect of miR-431 on CM proliferation, suggesting that FBXO32 is a functional target of miR-431 during CM proliferation. CONCLUSION: In summary, miR-431 promotes CM proliferation by targeting FBXO32, providing a potential molecular target for preventing myocardial injury.

A scalable federated learning solution for secondary care using low-cost microcomputing: privacy-preserving development and evaluation of a COVID-19 screening test in UK hospitals.

BACKGROUND: Multicentre training could reduce biases in medical artificial intelligence (AI); however, ethical, legal, and technical considerations can constrain the ability of hospitals to share data. Federated learning enables institutions to participate in algorithm development while retaining custody of their data but uptake in hospitals has been limited, possibly as deployment requires specialist software and technical expertise at each site. We previously developed an artificial intelligence-driven screening test for COVID-19 in emergency departments, known as CURIAL-Lab, which uses vital signs and blood tests that are routinely available within 1 h of a patient's arrival. Here we aimed to federate our COVID-19 screening test by developing an easy-to-use embedded system-which we introduce as full-stack federated learning-to train and evaluate machine learning models across four UK hospital groups without centralising patient data. METHODS: We supplied a Raspberry Pi 4 Model B preloaded with our federated learning software pipeline to four National Health Service (NHS) hospital groups in the UK: Oxford University Hospitals NHS Foundation Trust (OUH; through the locally linked research University, University of Oxford), University Hospitals Birmingham NHS Foundation Trust (UHB), Bedfordshire Hospitals NHS Foundation Trust (BH), and Portsmouth Hospitals University NHS Trust (PUH). OUH, PUH, and UHB participated in federated training, training a deep neural network and logistic regressor over 150 rounds to form and calibrate a global model to predict COVID-19 status, using clinical data from patients admitted before the pandemic (COVID-19-negative) and testing positive for COVID-19 during the first wave of the pandemic. We conducted a federated evaluation of the global model for admissions during the second wave of the pandemic at OUH, PUH, and externally at BH. For OUH and PUH, we additionally performed local fine-tuning of the global model using the sites' individual training data, forming a site-tuned model, and evaluated the resultant model for admissions during the second wave of the pandemic. This study included data collected between Dec 1, 2018, and March 1, 2021; the exact date ranges used varied by site. The primary outcome was overall model performance, measured as the area under the receiver operating characteristic curve (AUROC). Removable micro secure digital (microSD) storage was destroyed on study completion. FINDINGS: Clinical data from 130 941 patients (1772 COVID-19-positive), routinely collected across three hospital groups (OUH, PUH, and UHB), were included in federated training. The evaluation step included data from 32 986 patients (3549 COVID-19-positive) attending OUH, PUH, or BH during the second wave of the pandemic. Federated training of a global deep neural network classifier improved upon performance of models trained locally in terms of AUROC by a mean of 27·6% (SD 2·2): AUROC increased from 0·574 (95% CI 0·560-0·589) at OUH and 0·622 (0·608-0·637) at PUH using the locally trained models to 0·872 (0·862-0·882) at OUH and 0·876 (0·865-0·886) at PUH using the federated global model. Performance improvement was smaller for a logistic regression model, with a mean increase in AUROC of 13·9% (0·5%). During federated external evaluation at BH, AUROC for the global deep neural network model was 0·917 (0·893-0·942), with 89·7% sensitivity (83·6-93·6) and 76·6% specificity (73·9-79·1). Site-specific tuning of the global model did not significantly improve performance (change in AUROC <0·01). INTERPRETATION: We developed an embedded system for federated learning, using microcomputing to optimise for ease of deployment. We deployed full-stack federated learning across four UK hospital groups to develop a COVID-19 screening test without centralising patient data. Federation improved model performance, and the resultant global models were generalisable. Full-stack federated learning could enable hospitals to contribute to AI development at low cost and without specialist technical expertise at each site. FUNDING: The Wellcome Trust, University of Oxford Medical and Life Sciences Translational Fund.

The role of matrix metalloproteinase 9 in fibrosis diseases and its molecular mechanisms.

Fibrosis is a process of tissue repair that results in the slow creation of scar tissue to replace healthy tissue and can affect any tissue or organ. Its primary feature is the massive deposition of extracellular matrix (mainly collagen), eventually leading to tissue dysfunction and organ failure. The progression of fibrotic diseases has put a significant strain on global health and the economy, and as a result, there is an urgent need to find some new therapies. Previous studies have identified that inflammation, oxidative stress, some cytokines, and remodeling play a crucial role in fibrotic diseases and are essential avenues for treating fibrotic diseases. Among them, matrix metalloproteinases (MMPs) are considered the main targets for the treatment of fibrotic diseases since they are the primary driver involved in ECM degradation, and tissue inhibitors of metalloproteinases (TIMPs) are natural endogenous inhibitors of MMPs. Through previous studies, we found that MMP-9 is an essential target for treating fibrotic diseases. However, it is worth noting that MMP-9 plays a bidirectional regulatory role in different fibrotic diseases or different stages of the same fibrotic disease. Previously identified MMP-9 inhibitors, such as pirfenidone and nintedanib, suffer from some rather pronounced side effects, and therefore, there is an urgent need to investigate new drugs. In this review, we explore the mechanism of action and signaling pathways of MMP-9 in different tissues and organs, hoping to provide some ideas for developing safer and more effective biologics.

Student Loss: Towards the Probability Assumption in Inaccurate Supervision.

Noisy labels are often encountered in datasets, but learning with them is challenging. Although natural discrepancies between clean and mislabeled samples in a noisy category exist, most techniques in this field still gather them indiscriminately, which leads to their performances being partially robust. In this paper, we reveal both empirically and theoretically that the learning robustness can be improved by assuming deep features with the same labels follow a student distribution, resulting in a more intuitive method called student loss. By embedding the student distribution and exploiting the sharpness of its curve, our method is naturally data-selective and can offer extra strength to resist mislabeled samples. This ability makes clean samples aggregate tightly in the center, while mislabeled samples scatter, even if they share the same label. Additionally, we employ the metric learning strategy and develop a large-margin student (LT) loss for better capability. It should be noted that our approach is the first work that adopts the prior probability assumption in feature representation to decrease the contributions of mislabeled samples. This strategy can enhance various losses to join the student loss family, even if they have been robust losses. Experiments demonstrate that our approach is more effective in inaccurate supervision. Enhanced LT losses significantly outperform various state-of-the-art methods in most cases. Even huge improvements of over 50% can be obtained under some conditions. An implementation of the main codes is available at https://github.com/Zhangshuojackpot/Student-Loss.

Systemic inflammatory biomarkers in Schizophrenia are changed by ECT administration and related to the treatment efficacy.

Immune inflammation has long been implicated in the pathogenesis of schizophrenia. Despite as a rapid and effective physical therapy, the role of immune inflammation in electroconvulsive therapy (ECT) for schizophrenia remains elusive. The neutrophils to lymphocytes (NLR), platelets to monocytes (PLR) and monocytes to lymphocytes (MLR) are inexpensive and accessible biomarkers of systemic inflammation. In this study, 70 schizophrenia patients and 70 age- and sex-matched healthy controls were recruited. The systemic inflammatory biomarkers were measured before and after ECT. Our results indicated schizophrenia had significantly higher peripheral NLR, PLR and MLR compared to health controls at baseline, while lymphocytes did not differ. After 6 ECT, the psychiatric symptoms were significantly improved, as demonstrated by the Positive and Negative Syndrome Scale (PANSS). However, there was a decline in cognitive function scores, as indicated by the Mini-Mental State Examination (MMSE). Notably, the neutrophils and NLR were significantly reduced following ECT. Although lymphocytes remained unchanged following ECT, responders had significantly higher lymphocytes compared to non-responders. Moreover, the linear regression analyses revealed that higher lymphocytes served as a predictor of larger improvement in positive symptom following ECT. Overall, our findings further highlighted the presence of systemic inflammation in schizophrenia patients, and that ECT may exert a therapeutic effect in part by attenuating systemic inflammation. Further research may therefore lead to new treatment strategies for schizophrenia targeting the immune system.