Most urological conditions are represented by male disorders. Analyzing certain epidemiological aspects of female urological disorders at the National University Teaching Hospital of Cotnou, will allow us objectively ...Most urological conditions are represented by male disorders. Analyzing certain epidemiological aspects of female urological disorders at the National University Teaching Hospital of Cotnou, will allow us objectively to have reliable data to optimize the urological management of women. Patients and Methods: This was a retrospective, analytical and descriptive study that was conducted at the CNHU-HKM over a 10-year period from January 2008 to December 2017 on the epidemiological aspects of female urological diseases from the study of patient records. The variable studied was: the age, the profession, the year of admission, the organ affected, and the type of pathology. The confidentiality of the data has been ensured. Results: Female urological disorders over 10 years accounted for 9.62% of all patients received in Urology Department. The age group greater than 50 years was the most common at 31.34%. Traders and housewives were the most numerous (39.93% and 14.55%). Bladder diseases were the most frequent (51.50), dominated by vesico-vaginal fistulas (29.85%). In addition to fistulas, tumoral affections were most prevalent at 27.99% followed by infectious diseases at 8.21%. The lithiasic affections were infrequent at 6.72%. Conclusion: Female urological conditions are infrequent but not negligible, dominated by vesico-vaginal fistulas and tumors in the context of developing countries. The specific evaluation of each pathology group can help optimize management.展开更多
With the proliferation of Internet of Things(IoT)devices,securing these interconnected systems against cyberattacks has become a critical challenge.Traditional security paradigms often fail to cope with the scale and ...With the proliferation of Internet of Things(IoT)devices,securing these interconnected systems against cyberattacks has become a critical challenge.Traditional security paradigms often fail to cope with the scale and diversity of IoT network traffic.This paper presents a comparative benchmark of classic machine learning(ML)and state-of-the-art deep learning(DL)algorithms for IoT intrusion detection.Our methodology employs a twophased approach:a preliminary pilot study using a custom-generated dataset to establish baselines,followed by a comprehensive evaluation on the large-scale CICIoTDataset2023.We benchmarked algorithms including Random Forest,XGBoost,CNN,and StackedLSTM.The results indicate that while top-performingmodels frombothcategories achieve over 99%classification accuracy,this metric masks a crucial performance trade-off.We demonstrate that treebased ML ensembles exhibit superior precision(91%)in identifying benign traffic,making them effective at reducing false positives.Conversely,DL models demonstrate superior recall(96%),making them better suited for minimizing the interruption of legitimate traffic.We conclude that the selection of an optimal model is not merely a matter of maximizing accuracy but is a strategic choice dependent on the specific security priority either minimizing false alarms or ensuring service availability.Thiswork provides a practical framework for deploying context-aware security solutions in diverse IoT environments.展开更多
Climate change is impacting forests in Central Europe,causing increased mortality and degradation of forest ecosystem services(FES).As global warming intensifies,these effects are likely to worsen,particularly through...Climate change is impacting forests in Central Europe,causing increased mortality and degradation of forest ecosystem services(FES).As global warming intensifies,these effects are likely to worsen,particularly through more severe droughts and increased biotic disturbances.Understanding how forests respond to different levels of warming is essential for adaptation planning.Therefore,this study analyzed changes in forest structure and FES,including timber production,climate change mitigation,recreation,and structural diversity,under three global warming scenarios.Using the LandClim model,we compared warming levels of 1.5,2,and 3℃above preindustrial temperatures,based on 30-year periods from RCP data,to historical climate.Our research focused on Freiburg's forests in southwestern Germany,characterized by diverse tree species and an elevation range of 200–1,250 m a.s.l.A warming of 1.5℃could temporarily increase productivity,but at 2℃,biomass losses of up to 10%would occur below elevations of 450 m due to drought mortality.Under 3℃,losses would intensify below 650 m up to 40%,with even drought-resistant species like pedunculate oak experiencing mortality.At higher elevations,bark beetle outbreaks caused mortality of Norway spruce,while European beech capitalized on the changing ecological conditions.Higher warming levels significantly deteriorated FES,particularly timber production,climate change mitigation,and structural diversity,while recreation was less affected.These findings emphasize the urgency of meeting Paris Agreement targets,as limiting warming below 2℃can reduce severe impacts.If warming exceeds this critical threshold,even species presently considered drought-resistant,such as native sessile and pedunculate oaks and non-native red oak,could face serious threats at lower elevations.This would undermine the effectiveness of current management strategies,as these tree species are key to providing multiple FES.展开更多
Rapid evolutions of the Internet of Electric Vehicles(IoEVs)are reshaping and modernizing transport systems,yet challenges remain in energy efficiency,better battery aging,and grid stability.Typical charging methods a...Rapid evolutions of the Internet of Electric Vehicles(IoEVs)are reshaping and modernizing transport systems,yet challenges remain in energy efficiency,better battery aging,and grid stability.Typical charging methods allow for EVs to be charged without thought being given to the condition of the battery or the grid demand,thus increasing energy costs and battery aging.This study proposes a smart charging station with an AI-powered Battery Management System(BMS),developed and simulated in MATLAB/Simulink,to increase optimality in energy flow,battery health,and impractical scheduling within the IoEV environment.The system operates through real-time communication,load scheduling based on priorities,and adaptive charging based on batterymathematically computed State of Charge(SOC),State of Health(SOH),and thermal state,with bidirectional power flow(V2G),thus allowing EVs’participation towards grid stabilization.Simulation results revealed that the proposed model can reduce peak grid load by 37.8%;charging efficiency is enhanced by 92.6%;battery temperature lessened by 4.4℃;SOH extended over 100 cycles by 6.5%,if compared against the conventional technique.By this way,charging time was decreased by 12.4% and energy costs dropped by more than 20%.These results showed that smart charging with intelligent BMS can boost greatly the operational efficiency and sustainability of the IoEV ecosystem.展开更多
Xylogenesis,the process through which wood cells are formed,results in the long-term storage of carbon in woody biomass,making it a key component of the global carbon cycle.Understanding how environmental drivers infl...Xylogenesis,the process through which wood cells are formed,results in the long-term storage of carbon in woody biomass,making it a key component of the global carbon cycle.Understanding how environmental drivers influence xylogenesis during the growing season is therefore of great interest.However,studying shortterm drivers of wood production using xylogenetic data is complicated by the usual sampling scheme and the influence of eccentric growth,i.e.,heterogeneous growth around the stem.In this study,we improve xylogenesis research by introducing a statistical approach that explicitly considers seasonal phenology,short-term growth rates,and growth eccentricity.To this end,we developed Bayesian models of xylogenesis and compared them with a conventional method based on the use of Gompertz functions.Our results show that eccentricity generated high temporal autocorrelation between successive samples,and that explicitly taking it into account improved both the representativeness of phenology and intra-ring variability.We observed consistent short-term patterns in the model residuals,suggesting the influence of an unaccounted-for environmental variable on cell production.The proposed models offer several advantages over traditional methods,including robust confidence intervals around predictions,consistency with phenology,and reduced sensitivity to extreme observations at the end of the growing season,often linked to eccentric growth.These models also provide a benchmark for mechanistic testing of short-term drivers of wood formation.展开更多
Titanium exhibits outstanding properties,particularly,high specific strength and resistance to both high and low temperatures,earning it a reputation as the metal of the future.However,because of the highly reactive n...Titanium exhibits outstanding properties,particularly,high specific strength and resistance to both high and low temperatures,earning it a reputation as the metal of the future.However,because of the highly reactive nature of titanium,metallic titanium production involves extensive procedures and high costs.Considering its advantages and limitations,the European Union has classified titanium metal as a critical raw material(CRM)of low category.The Kroll process is predominantly used to produce titanium;however,molten salt electrolysis(MSE)is currently being explored for producing metallic titanium at a low cost.Since 2000,electrolytic titanium production has undergone a wave of technological advancements.However,because of the intermediate and disproportionation reactions in the electrolytic titanium production process,the process efficiency and titanium purity according to industrial standards could not be achieved.Consequently,metallic titanium production has gradually diversified into employing technologies such as thermal reduction,MSE,and titanium alloy preparation.This study provides a comprehensive review of research advances in titanium metal preparation technologies over the past two decades,highlighting the challenges faced by the existing methods and proposing potential solutions.It offers useful insights into the development of low-cost titanium preparation technologies.展开更多
Background:Stomach cancer(SC)is one of the most lethal malignancies worldwide due to late-stage diagnosis and limited treatment.The transcriptomic,epigenomic,and proteomic,etc.,omics datasets generated by high-through...Background:Stomach cancer(SC)is one of the most lethal malignancies worldwide due to late-stage diagnosis and limited treatment.The transcriptomic,epigenomic,and proteomic,etc.,omics datasets generated by high-throughput sequencing technology have become prominent in biomedical research,and they reveal molecular aspects of cancer diagnosis and therapy.Despite the development of advanced sequencing technology,the presence of high-dimensionality in multi-omics data makes it challenging to interpret the data.Methods:In this study,we introduce RankXLAN,an explainable ensemble-based multi-omics framework that integrates feature selection(FS),ensemble learning,bioinformatics,and in-silico validation for robust biomarker detection,potential therapeutic drug-repurposing candidates’identification,and classification of SC.To enhance the interpretability of the model,we incorporated explainable artificial intelligence(SHapley Additive exPlanations analysis),as well as accuracy,precision,F1-score,recall,cross-validation,specificity,likelihood ratio(LR)+,LR−,and Youden index results.Results:The experimental results showed that the top four FS algorithms achieved improved results when applied to the ensemble learning classification model.The proposed ensemble model produced an area under the curve(AUC)score of 0.994 for gene expression,0.97 for methylation,and 0.96 for miRNA expression data.Through the integration of bioinformatics and ML approach of the transcriptomic and epigenomic multi-omics dataset,we identified potential marker genes,namely,UBE2D2,HPCAL4,IGHA1,DPT,and FN3K.In-silico molecular docking revealed a strong binding affinity between ANKRD13C and the FDA-approved drug Everolimus(binding affinity−10.1 kcal/mol),identifying ANKRD13C as a potential therapeutic drug-repurposing target for SC.Conclusion:The proposed framework RankXLAN outperforms other existing frameworks for serum biomarker identification,therapeutic target identification,and SC classification with multi-omics datasets.展开更多
Dysphagia,a prevalent condition affecting over 30%of the elderly,significantly elevates malnutrition risks due to impaired swallowing and insufficient nutrient intake.This study aimed to develop plant-based,3D-printed...Dysphagia,a prevalent condition affecting over 30%of the elderly,significantly elevates malnutrition risks due to impaired swallowing and insufficient nutrient intake.This study aimed to develop plant-based,3D-printed dysphagia diets using pea protein isolate(PPI)combined with quinoa to enhance essential amino acid profiles,complemented by hydrocolloid—xanthan gum(XG),carboxymethyl cellulose(CMC),and agar—for tailored texture modulation.Eight ink formulations were evaluated based on molecular interactions,rheological behavior,3 D printing performance,and compliance with International Dysphagia Diet S tandardization Initiative(IDDSI)standards.Synergistic effects of XG and CMC in Ink-C optimized shear-thinning properties and structural stability,enabling high-precision printing of self-supporting constructs.IDDSI testing confirmed that Ink-A and Ink-C met Level 5"minced and moist"criteria,validated by texture parameters and shape retention during mechanical testing.Electronic nose showed minimal deviations in aromatic characteristics across all formulations,preserving sensory acceptability.In vitro digestion models revealed that hydrocolloid networks temporarily hindered gastric proteolysis but ultimately achieved sufficient intestinal hydrolysis(>76%)to ensure nutrient bioavailability.Ink-C was identified as the optimal formulation,harmonizing printability,swallow-safe textures,and digestibility.This work highlights the potential of hydrocolloid-engineered 3D printing to advance personalized nutrition for dysphagia management,offering scalable solutions to improve dietary diversity and clinical outcomes in aging populations.展开更多
This review comprehensively summarized the potential of artificial intelligence(AI)in the management of esophageal cancer.It highlighted the significance of AI-assisted endoscopy in Japan where endoscopy is central to...This review comprehensively summarized the potential of artificial intelligence(AI)in the management of esophageal cancer.It highlighted the significance of AI-assisted endoscopy in Japan where endoscopy is central to both screening and diagnosis.For the clinical adaptation of AI,several challenges remain for its effective translation.The establishment of high-quality clinical databases,such as the National Clinical Database and Japan Endoscopy Database in Japan,which covers almost all cases of esophageal cancer,is essential for validating multimodal AI models.This requires rigorous external validation using diverse datasets,including those from different endoscope manufacturers and image qualities.Furthermore,endoscopists’skills significantly affect diagnostic accuracy,suggesting that AI should serve as a supportive tool rather than a replacement.Addressing these challenges,along with country-specific legal and ethical considerations,will facilitate the successful integration of multimodal AI into the management of esophageal cancer,particularly in endoscopic diagnosis,and contribute to improved patient outcomes.Although this review focused on Japan as a case study,the challenges and solutions described are broadly applicable to other high-incidence regions.展开更多
This review provides a comprehensive overview of recent advancements in aluminum-based conductor alloys engineered to achieve superior mechanical strength and thermal stability without sacrificing electrical conductiv...This review provides a comprehensive overview of recent advancements in aluminum-based conductor alloys engineered to achieve superior mechanical strength and thermal stability without sacrificing electrical conductivity.Particular emphasis is placed on the role of microalloying elements—particularly Sc and Zr-in promoting the formation of coherent nanoscale precipitates such as Al_(3)Zr,Al_(3)Sc,and core-shell Al_(3)(Sc,Zr)with metastable L1_(2)crystal structures.These precipitates contribute significantly to high-temperature performance by enabling precipitation strengthening and stabilizing grain boundaries.The review also explores the emerging role of other rare earth elements(REEs),such as erbium(Er),in accelerating precipitation kinetics and improving thermal stability by retarding coarsening.Additionally,recent advancements in thermomechanical processing strategies are examined,with a focus on scalable approaches to optimize the strength-conductivity balance.These approaches involve multi-step heat treatments and carefully controlled manufacturing sequences,particularly the combination of cold drawing and aging treatment to promote uniform and effective precipitation.This review offers valuable insights to guide the development of cost-effective,high-strength,heat-resistant aluminum alloys beyond conductor applications,particularly those strengthened through microalloying with Sc and Zr.展开更多
Background:That Central and Eastern Europe and Central Asia(CEECA)experienced a major mortality crisis in the 1990s is a well-established finding,with most analyses focusing on singular causes like alcohol-related dea...Background:That Central and Eastern Europe and Central Asia(CEECA)experienced a major mortality crisis in the 1990s is a well-established finding,with most analyses focusing on singular causes like alcohol-related deaths.However,the utility of the integrated“deaths of despair”framework,which views alcohol,drug,and suicide deaths as a unified socio-economic phenomenon,remains under-explored in this context.Crucially,the long-term evolution of the composition of despair within the region remains a largely unexplored area of inquiry.Therefore,this study aims to analyze the long-term trends,changing composition,and regional heterogeneity of deaths from despair in the CEECA region from 1980 to 2021.Methods:Using 2021 Global Burden of Disease(GBD)data(1980–2021),we analyzed deaths of despair mortality trends in 29 CEECA countries.We employed Joinpoint regression to identify significant trend changes and conducted stratified analyses by cause,gender,and age group.Results:The CEECA deaths of despair crisis began as an alcohol and suicide driven phenomenon concentrated in middle-aged men(50–74 years)during the 1990s,with mortality rates for alcohol use disorders and self-harm surging annually by 30.35%(p=0.002)and 13.44%(p=0.001),respectively,between 1991 and 1994.It has since evolved,marked by a contrasting and emerging threat in the 21st century:a rising proportion of drug-related deaths among the younger(15–49 years)male cohort,where the share of drug use disorders increased from 6.9%in 2000 to 11.8%in 2008.Conclusion:The deaths of despair crisis in the CEECA region is not a past event but an ongoing,evolving phenomenon.Its changing nature demands a shift in public health focus from solely historical drivers to new,generation-specific threats,particularly the rise of drug-related despair among youth.展开更多
Interferon-related genes are involved in antiviral responses,inflammation,and immunity,which are closely related to sepsis-associated acute respiratory distress syndrome(ARDS).We analyzed 1972 participants with genoty...Interferon-related genes are involved in antiviral responses,inflammation,and immunity,which are closely related to sepsis-associated acute respiratory distress syndrome(ARDS).We analyzed 1972 participants with genotype data and 681 participants with gene expression data from the Molecular Epidemiology of ARDS(MEARDS),the Molecular Epidemiology of Sepsis in the ICU(MESSI),and the Molecular Diagnosis and Risk Stratification of Sepsis(MARS)cohorts in a three-step study focusing on sepsis-associated ARDS and sepsis-only controls.First,we identified and validated interferon-related genes associated with sepsis-associated ARDS risk using genetically regulated gene expression(GReX).Second,we examined the association of the confirmed gene(interferon regulatory factor 1,IRF1)with ARDS risk and survival and conducted a mediation analysis.Through discovery and validation,we found that the GReX of IRF1 was associated with ARDS risk(odds ratio[OR_(MEARDS)]=0.84,P=0.008;OR_(MESSI)=0.83,P=0.034).Furthermore,individual-level measured IRF1 expression was associated with reduced ARDS risk(OR=0.58,P=8.67×10^(-4)),and improved overall survival in ARDS patients(hazard ratio[HR_(28-day)]=0.49,P=0.009)and sepsis patients(HR_(28-day)=0.76,P=0.008).Mediation analysis revealed that IRF1 may enhance immune function by regulating the major histocompatibility complex,including HLA-F,which mediated more than 70%of protective effects of IRF1 on ARDS.The findings were validated by in vitro biological experiments including time-series infection dynamics,overexpression,knockout,and chromatin immunoprecipitation sequencing.Early prophylactic interventions to activate IRF1 in sepsis patients,thereby regulating HLA-F,may reduce the risk of ARDS and mortality,especially in severely ill patients.展开更多
Deep neural networks have achieved excellent classification results on several computer vision benchmarks.This has led to the popularity of machine learning as a service,where trained algorithms are hosted on the clou...Deep neural networks have achieved excellent classification results on several computer vision benchmarks.This has led to the popularity of machine learning as a service,where trained algorithms are hosted on the cloud and inference can be obtained on real-world data.In most applications,it is important to compress the vision data due to the enormous bandwidth and memory requirements.Video codecs exploit spatial and temporal correlations to achieve high compression ratios,but they are computationally expensive.This work computes the motion fields between consecutive frames to facilitate the efficient classification of videos.However,contrary to the normal practice of reconstructing the full-resolution frames through motion compensation,this work proposes to infer the class label from the block-based computed motion fields directly.Motion fields are a richer and more complex representation of motion vectors,where each motion vector carries the magnitude and direction information.This approach has two advantages:the cost of motion compensation and video decoding is avoided,and the dimensions of the input signal are highly reduced.This results in a shallower network for classification.The neural network can be trained using motion vectors in two ways:complex representations and magnitude-direction pairs.The proposed work trains a convolutional neural network on the direction and magnitude tensors of the motion fields.Our experimental results show 20×faster convergence during training,reduced overfitting,and accelerated inference on a hand gesture recognition dataset compared to full-resolution and downsampled frames.We validate the proposed methodology on the HGds dataset,achieving a testing accuracy of 99.21%,on the HMDB51 dataset,achieving 82.54%accuracy,and on the UCF101 dataset,achieving 97.13%accuracy,outperforming state-of-the-art methods in computational efficiency.展开更多
This study investigates the performance enhancement of super-sulfated cement(SSC)derived from arsenic-containing bio-oxidation waste(BW)through the incorporation of carbonated recycled concrete fines(CRCF).The finding...This study investigates the performance enhancement of super-sulfated cement(SSC)derived from arsenic-containing bio-oxidation waste(BW)through the incorporation of carbonated recycled concrete fines(CRCF).The findings revealed that the addition of 5wt%CRCF yields optimal performance,with compressive strengths reaching approximately 1.83,12.59,and 42.81 MPa at 1,3,and 28 d,respectively.These values represented significant increases of 408.3%,10.0%,and 14.3%compared to the reference sample.The improvement was attributed to the synergistic effects of ultrafine CRCF particles acting as fillers and nucleation sites,as well as the high reactivity of silica gels,which promoted the formation of additional hydration gels.Microstructural analysis confirmed that CRCF addition refined pore structure,and enhanced the stiffness of C-S-H gels.Furthermore,CRCF served as a net CO_(2) sink,sequestering 0.268 kg CO_(2) per kilogram of CRCF and thereby reducing the carbon footprint of SSC.In addition,the feasibility of applying CRCF-modified SSC in cemented paste backfill(CPB)is highlighted,given the high cement-related carbon footprint of conventional CPB.When 5wt%CRCFmodified SSC was employed in CPB,its 3-d compressive strength attained over 70%of that of ordinary Portland cement(OPC),while the 28-d strength was comparable to that of OPC.The proposed binder thus provides a sustainable pathway for BW valorization,combining waste utilization,carbon sequestration,and improved engineering performance.展开更多
Background:Investigators from low-,middle-,and high-income countries representing 6 continents contributed to the development of the Global Adolescent and Child Physical Activity Questionnaire(GAC-PAQ).The GAC-PAQ is ...Background:Investigators from low-,middle-,and high-income countries representing 6 continents contributed to the development of the Global Adolescent and Child Physical Activity Questionnaire(GAC-PAQ).The GAC-PAQ is designed to assess physical activity(PA)across all key domains(i.e.,school,chores,work/volunteering,transport,free time,outdoor time).It aimed to address multiple gaps in global PA surveillance(e.g.,omission of important PA domains,insufficient cultural adaptation,underrepresentation of rural areas in questionnaire validation studies).The purpose of this study was to assess the content validity of the GAC-PAQ among PA experts,8-to 17-year-olds,and one of their parents/guardians,and to discuss changes made to the questionnaire based on participants'feedback.Methods:Sixty-two experts in PA measurement and/or surveillance from 24 countries completed an online survey that included both closed-and open-ended questions about the content validity of the GAC-PAQ.The proportion of experts who agreed or strongly agreed with the items was calculated.Child-parent/guardian dyads from 15 countries(n=250;10-40 per country)participated in a structured cognitive interview to assess the clarity of the questions and response options,and they were encouraged to provide suggestions to improve clarity and facilitate completion of the questionnaire.Participating countries are:Aotearoa New Zealand,Brazil,Canada,China,Colombia,Czech Republic,India,Malawi,Mexico,Nepal,Nigeria,Spain,Sweden,Thailand,and the United Arab Emirates.Interviews were conducted in 13 different languages and structured by PA domain.Generic images were included to help participants in answering questions about PA intensity.Results:Expert agreement with the items for each domain exceeded 75%,and their qualitative feedback was used to revise the questionnaire before cognitive interviews.In general,participants found the questionnaire to be comprehensive.Adolescents(12-17 years)found it easier than children(8-11 years)to answer the questions.Several children struggled to answer questions about the duration and intensity of activities and/or concepts related to travel modes,active trips,and organized activities.Many parents/guardians were unsure about the frequency,duration,and intensity of their children's or adolescents'PA at school and/or recommended using more culturally relevant and appropriate images.Some participants misunderstood the concept of activities that“make you stronger”(intended to assess resistance activities)and/or struggled to differentiate between work,volunteering,and chores.Conclusion:Participants'feedback was used to develop a revised,simplified,and culturally adapted GAC-PAQ,which will be pilot-tested in all15 countries in an App that will include country-specific images and narration in local languages.Further research is needed to assess the reliability and validity of the revised GAC-PAQ.展开更多
At inference time,deep neural networks are susceptible to backdoor attacks,which can produce attackercontrolled outputs when inputs contain carefully crafted triggers.Existing defense methods often focus on specific a...At inference time,deep neural networks are susceptible to backdoor attacks,which can produce attackercontrolled outputs when inputs contain carefully crafted triggers.Existing defense methods often focus on specific attack types or incur high costs,such as data cleaning or model fine-tuning.In contrast,we argue that it is possible to achieve effective and generalizable defense without removing triggers or incurring high model-cleaning costs.Fromthe attacker’s perspective and based on characteristics of vulnerable neuron activation anomalies,we propose an Adaptive Feature Injection(AFI)method for black-box backdoor detection.AFI employs a pre-trained image encoder to extract multi-level deep features and constructs a dynamic weight fusionmechanism for precise identification and interception of poisoned samples.Specifically,we select the control samples with the largest feature differences fromthe clean dataset via feature-space analysis,and generate blended sample pairs with the test sample using dynamic linear interpolation.The detection statistic is computed by measuring the divergence G(x)in model output responses.We systematically evaluate the effectiveness of AFI against representative backdoor attacks,including BadNets,Blend,WaNet,and IAB,on three benchmark datasets:MNIST,CIFAR-10,and ImageNet.Experimental results show that AFI can effectively detect poisoned samples,achieving average detection rates of 95.20%,94.15%,and 86.49%on these datasets,respectively.Compared with existing methods,AFI demonstrates strong cross-domain generalization ability and robustness to unknown attacks.展开更多
The rapid advancement of machine learning based tight-binding Hamiltonian(MLTB)methods has opened new avenues for efficient and accurate electronic structure simulations,particularly in large-scale systems and long-ti...The rapid advancement of machine learning based tight-binding Hamiltonian(MLTB)methods has opened new avenues for efficient and accurate electronic structure simulations,particularly in large-scale systems and long-time scenarios.This review begins with a concise overview of traditional tight-binding(TB)models,including both(semi-)empirical and first-principles approaches,establishing the foundation for understanding MLTB developments.We then present a systematic classification of existing MLTB methodologies,grouped into two major categories:direct prediction of TB Hamiltonian elements and inference of empirical parameters.A comparative analysis with other ML-based electronic structure models is also provided,highlighting the advancement of MLTB approaches.Finally,we explore the emerging MLTB application ecosystem,highlighting how the integration of MLTB models with a diverse suite of post-processing tools from linear-scaling solvers to quantum transport frameworks and molecular dynamics interfaces is essential for tackling complex scientific problems across different domains.The continued advancement of this integrated paradigm promises to accelerate materials discovery and open new frontiers in the predictive simulation of complex quantum phenomena.展开更多
文摘Most urological conditions are represented by male disorders. Analyzing certain epidemiological aspects of female urological disorders at the National University Teaching Hospital of Cotnou, will allow us objectively to have reliable data to optimize the urological management of women. Patients and Methods: This was a retrospective, analytical and descriptive study that was conducted at the CNHU-HKM over a 10-year period from January 2008 to December 2017 on the epidemiological aspects of female urological diseases from the study of patient records. The variable studied was: the age, the profession, the year of admission, the organ affected, and the type of pathology. The confidentiality of the data has been ensured. Results: Female urological disorders over 10 years accounted for 9.62% of all patients received in Urology Department. The age group greater than 50 years was the most common at 31.34%. Traders and housewives were the most numerous (39.93% and 14.55%). Bladder diseases were the most frequent (51.50), dominated by vesico-vaginal fistulas (29.85%). In addition to fistulas, tumoral affections were most prevalent at 27.99% followed by infectious diseases at 8.21%. The lithiasic affections were infrequent at 6.72%. Conclusion: Female urological conditions are infrequent but not negligible, dominated by vesico-vaginal fistulas and tumors in the context of developing countries. The specific evaluation of each pathology group can help optimize management.
文摘With the proliferation of Internet of Things(IoT)devices,securing these interconnected systems against cyberattacks has become a critical challenge.Traditional security paradigms often fail to cope with the scale and diversity of IoT network traffic.This paper presents a comparative benchmark of classic machine learning(ML)and state-of-the-art deep learning(DL)algorithms for IoT intrusion detection.Our methodology employs a twophased approach:a preliminary pilot study using a custom-generated dataset to establish baselines,followed by a comprehensive evaluation on the large-scale CICIoTDataset2023.We benchmarked algorithms including Random Forest,XGBoost,CNN,and StackedLSTM.The results indicate that while top-performingmodels frombothcategories achieve over 99%classification accuracy,this metric masks a crucial performance trade-off.We demonstrate that treebased ML ensembles exhibit superior precision(91%)in identifying benign traffic,making them effective at reducing false positives.Conversely,DL models demonstrate superior recall(96%),making them better suited for minimizing the interruption of legitimate traffic.We conclude that the selection of an optimal model is not merely a matter of maximizing accuracy but is a strategic choice dependent on the specific security priority either minimizing false alarms or ensuring service availability.Thiswork provides a practical framework for deploying context-aware security solutions in diverse IoT environments.
基金funded by the HORIZON EUROPE's project"eco2adapt"(Ecosystem-based Adaptation and Changemaking to Shape,Project,and Sustain the Resilience of Tomorrow's Forests,Grant no:101059498)。
文摘Climate change is impacting forests in Central Europe,causing increased mortality and degradation of forest ecosystem services(FES).As global warming intensifies,these effects are likely to worsen,particularly through more severe droughts and increased biotic disturbances.Understanding how forests respond to different levels of warming is essential for adaptation planning.Therefore,this study analyzed changes in forest structure and FES,including timber production,climate change mitigation,recreation,and structural diversity,under three global warming scenarios.Using the LandClim model,we compared warming levels of 1.5,2,and 3℃above preindustrial temperatures,based on 30-year periods from RCP data,to historical climate.Our research focused on Freiburg's forests in southwestern Germany,characterized by diverse tree species and an elevation range of 200–1,250 m a.s.l.A warming of 1.5℃could temporarily increase productivity,but at 2℃,biomass losses of up to 10%would occur below elevations of 450 m due to drought mortality.Under 3℃,losses would intensify below 650 m up to 40%,with even drought-resistant species like pedunculate oak experiencing mortality.At higher elevations,bark beetle outbreaks caused mortality of Norway spruce,while European beech capitalized on the changing ecological conditions.Higher warming levels significantly deteriorated FES,particularly timber production,climate change mitigation,and structural diversity,while recreation was less affected.These findings emphasize the urgency of meeting Paris Agreement targets,as limiting warming below 2℃can reduce severe impacts.If warming exceeds this critical threshold,even species presently considered drought-resistant,such as native sessile and pedunculate oaks and non-native red oak,could face serious threats at lower elevations.This would undermine the effectiveness of current management strategies,as these tree species are key to providing multiple FES.
文摘Rapid evolutions of the Internet of Electric Vehicles(IoEVs)are reshaping and modernizing transport systems,yet challenges remain in energy efficiency,better battery aging,and grid stability.Typical charging methods allow for EVs to be charged without thought being given to the condition of the battery or the grid demand,thus increasing energy costs and battery aging.This study proposes a smart charging station with an AI-powered Battery Management System(BMS),developed and simulated in MATLAB/Simulink,to increase optimality in energy flow,battery health,and impractical scheduling within the IoEV environment.The system operates through real-time communication,load scheduling based on priorities,and adaptive charging based on batterymathematically computed State of Charge(SOC),State of Health(SOH),and thermal state,with bidirectional power flow(V2G),thus allowing EVs’participation towards grid stabilization.Simulation results revealed that the proposed model can reduce peak grid load by 37.8%;charging efficiency is enhanced by 92.6%;battery temperature lessened by 4.4℃;SOH extended over 100 cycles by 6.5%,if compared against the conventional technique.By this way,charging time was decreased by 12.4% and energy costs dropped by more than 20%.These results showed that smart charging with intelligent BMS can boost greatly the operational efficiency and sustainability of the IoEV ecosystem.
基金supported by the Discovery Grants program of the Natural Sciences and Engineering Research Council of Canada(No.RGPIN-2021-03553)the Canadian Research Chair in dendroecology and dendroclimatology(CRC-2021-00368)+3 种基金the Ministère des Ressources Naturelles et des Forèts(MRNF,Contract no.142332177-D)the Natural Sciences and Engineering Research Council of Canada(Alliance Grant No.ALLRP 557148-20,obtained in partnership with the MRNF and Resolute Forest Products)the Fonds de recherche du Qu ebec–Nature et technologies(Partnership Research Program on the Contribution of the Forestry Sector to Climate Change MitigationGrant No.2022-0FC-309064)。
文摘Xylogenesis,the process through which wood cells are formed,results in the long-term storage of carbon in woody biomass,making it a key component of the global carbon cycle.Understanding how environmental drivers influence xylogenesis during the growing season is therefore of great interest.However,studying shortterm drivers of wood production using xylogenetic data is complicated by the usual sampling scheme and the influence of eccentric growth,i.e.,heterogeneous growth around the stem.In this study,we improve xylogenesis research by introducing a statistical approach that explicitly considers seasonal phenology,short-term growth rates,and growth eccentricity.To this end,we developed Bayesian models of xylogenesis and compared them with a conventional method based on the use of Gompertz functions.Our results show that eccentricity generated high temporal autocorrelation between successive samples,and that explicitly taking it into account improved both the representativeness of phenology and intra-ring variability.We observed consistent short-term patterns in the model residuals,suggesting the influence of an unaccounted-for environmental variable on cell production.The proposed models offer several advantages over traditional methods,including robust confidence intervals around predictions,consistency with phenology,and reduced sensitivity to extreme observations at the end of the growing season,often linked to eccentric growth.These models also provide a benchmark for mechanistic testing of short-term drivers of wood formation.
基金financial support from the Yunnan Province Key Industries Science and Technology Special Project for Colleges and UniversitiesChina(No.FWCY-QYCT2024006)+6 种基金National Natural Science Foundation of China(Nos.52104351 and 52364051)Science and Technology Major Project of Yunnan Province,China(No.202202AG050007)the Yunnan Fundamental Research ProjectsChina(No.202401AT070314)the Key Technology Research and Development Program of Shandong Province,China(No.2023CXGC010903)Central Guidance Local Scientific and Technological Development Funds,China(No.202407AB110022)Yunnan Province Xingdian Talent Support Plan Project,China。
文摘Titanium exhibits outstanding properties,particularly,high specific strength and resistance to both high and low temperatures,earning it a reputation as the metal of the future.However,because of the highly reactive nature of titanium,metallic titanium production involves extensive procedures and high costs.Considering its advantages and limitations,the European Union has classified titanium metal as a critical raw material(CRM)of low category.The Kroll process is predominantly used to produce titanium;however,molten salt electrolysis(MSE)is currently being explored for producing metallic titanium at a low cost.Since 2000,electrolytic titanium production has undergone a wave of technological advancements.However,because of the intermediate and disproportionation reactions in the electrolytic titanium production process,the process efficiency and titanium purity according to industrial standards could not be achieved.Consequently,metallic titanium production has gradually diversified into employing technologies such as thermal reduction,MSE,and titanium alloy preparation.This study provides a comprehensive review of research advances in titanium metal preparation technologies over the past two decades,highlighting the challenges faced by the existing methods and proposing potential solutions.It offers useful insights into the development of low-cost titanium preparation technologies.
基金the Deanship of Research and Graduate Studies at King Khalid University,KSA,for funding this work through the Large Research Project under grant number RGP2/164/46.
文摘Background:Stomach cancer(SC)is one of the most lethal malignancies worldwide due to late-stage diagnosis and limited treatment.The transcriptomic,epigenomic,and proteomic,etc.,omics datasets generated by high-throughput sequencing technology have become prominent in biomedical research,and they reveal molecular aspects of cancer diagnosis and therapy.Despite the development of advanced sequencing technology,the presence of high-dimensionality in multi-omics data makes it challenging to interpret the data.Methods:In this study,we introduce RankXLAN,an explainable ensemble-based multi-omics framework that integrates feature selection(FS),ensemble learning,bioinformatics,and in-silico validation for robust biomarker detection,potential therapeutic drug-repurposing candidates’identification,and classification of SC.To enhance the interpretability of the model,we incorporated explainable artificial intelligence(SHapley Additive exPlanations analysis),as well as accuracy,precision,F1-score,recall,cross-validation,specificity,likelihood ratio(LR)+,LR−,and Youden index results.Results:The experimental results showed that the top four FS algorithms achieved improved results when applied to the ensemble learning classification model.The proposed ensemble model produced an area under the curve(AUC)score of 0.994 for gene expression,0.97 for methylation,and 0.96 for miRNA expression data.Through the integration of bioinformatics and ML approach of the transcriptomic and epigenomic multi-omics dataset,we identified potential marker genes,namely,UBE2D2,HPCAL4,IGHA1,DPT,and FN3K.In-silico molecular docking revealed a strong binding affinity between ANKRD13C and the FDA-approved drug Everolimus(binding affinity−10.1 kcal/mol),identifying ANKRD13C as a potential therapeutic drug-repurposing target for SC.Conclusion:The proposed framework RankXLAN outperforms other existing frameworks for serum biomarker identification,therapeutic target identification,and SC classification with multi-omics datasets.
基金the National Key Research and Development Program of China(2022YFF1102400)for supporting this work。
文摘Dysphagia,a prevalent condition affecting over 30%of the elderly,significantly elevates malnutrition risks due to impaired swallowing and insufficient nutrient intake.This study aimed to develop plant-based,3D-printed dysphagia diets using pea protein isolate(PPI)combined with quinoa to enhance essential amino acid profiles,complemented by hydrocolloid—xanthan gum(XG),carboxymethyl cellulose(CMC),and agar—for tailored texture modulation.Eight ink formulations were evaluated based on molecular interactions,rheological behavior,3 D printing performance,and compliance with International Dysphagia Diet S tandardization Initiative(IDDSI)standards.Synergistic effects of XG and CMC in Ink-C optimized shear-thinning properties and structural stability,enabling high-precision printing of self-supporting constructs.IDDSI testing confirmed that Ink-A and Ink-C met Level 5"minced and moist"criteria,validated by texture parameters and shape retention during mechanical testing.Electronic nose showed minimal deviations in aromatic characteristics across all formulations,preserving sensory acceptability.In vitro digestion models revealed that hydrocolloid networks temporarily hindered gastric proteolysis but ultimately achieved sufficient intestinal hydrolysis(>76%)to ensure nutrient bioavailability.Ink-C was identified as the optimal formulation,harmonizing printability,swallow-safe textures,and digestibility.This work highlights the potential of hydrocolloid-engineered 3D printing to advance personalized nutrition for dysphagia management,offering scalable solutions to improve dietary diversity and clinical outcomes in aging populations.
基金Supported by Japan Society for the Promotion of Science,No.24K11935.
文摘This review comprehensively summarized the potential of artificial intelligence(AI)in the management of esophageal cancer.It highlighted the significance of AI-assisted endoscopy in Japan where endoscopy is central to both screening and diagnosis.For the clinical adaptation of AI,several challenges remain for its effective translation.The establishment of high-quality clinical databases,such as the National Clinical Database and Japan Endoscopy Database in Japan,which covers almost all cases of esophageal cancer,is essential for validating multimodal AI models.This requires rigorous external validation using diverse datasets,including those from different endoscope manufacturers and image qualities.Furthermore,endoscopists’skills significantly affect diagnostic accuracy,suggesting that AI should serve as a supportive tool rather than a replacement.Addressing these challenges,along with country-specific legal and ethical considerations,will facilitate the successful integration of multimodal AI into the management of esophageal cancer,particularly in endoscopic diagnosis,and contribute to improved patient outcomes.Although this review focused on Japan as a case study,the challenges and solutions described are broadly applicable to other high-incidence regions.
文摘This review provides a comprehensive overview of recent advancements in aluminum-based conductor alloys engineered to achieve superior mechanical strength and thermal stability without sacrificing electrical conductivity.Particular emphasis is placed on the role of microalloying elements—particularly Sc and Zr-in promoting the formation of coherent nanoscale precipitates such as Al_(3)Zr,Al_(3)Sc,and core-shell Al_(3)(Sc,Zr)with metastable L1_(2)crystal structures.These precipitates contribute significantly to high-temperature performance by enabling precipitation strengthening and stabilizing grain boundaries.The review also explores the emerging role of other rare earth elements(REEs),such as erbium(Er),in accelerating precipitation kinetics and improving thermal stability by retarding coarsening.Additionally,recent advancements in thermomechanical processing strategies are examined,with a focus on scalable approaches to optimize the strength-conductivity balance.These approaches involve multi-step heat treatments and carefully controlled manufacturing sequences,particularly the combination of cold drawing and aging treatment to promote uniform and effective precipitation.This review offers valuable insights to guide the development of cost-effective,high-strength,heat-resistant aluminum alloys beyond conductor applications,particularly those strengthened through microalloying with Sc and Zr.
基金supported by grants from the National Research Foundation of Korea(NRF)under the Ministry of Science and Information and Communication Technology(grant number:RS-2023-00249082)Korea University(grant number:K2225791).
文摘Background:That Central and Eastern Europe and Central Asia(CEECA)experienced a major mortality crisis in the 1990s is a well-established finding,with most analyses focusing on singular causes like alcohol-related deaths.However,the utility of the integrated“deaths of despair”framework,which views alcohol,drug,and suicide deaths as a unified socio-economic phenomenon,remains under-explored in this context.Crucially,the long-term evolution of the composition of despair within the region remains a largely unexplored area of inquiry.Therefore,this study aims to analyze the long-term trends,changing composition,and regional heterogeneity of deaths from despair in the CEECA region from 1980 to 2021.Methods:Using 2021 Global Burden of Disease(GBD)data(1980–2021),we analyzed deaths of despair mortality trends in 29 CEECA countries.We employed Joinpoint regression to identify significant trend changes and conducted stratified analyses by cause,gender,and age group.Results:The CEECA deaths of despair crisis began as an alcohol and suicide driven phenomenon concentrated in middle-aged men(50–74 years)during the 1990s,with mortality rates for alcohol use disorders and self-harm surging annually by 30.35%(p=0.002)and 13.44%(p=0.001),respectively,between 1991 and 1994.It has since evolved,marked by a contrasting and emerging threat in the 21st century:a rising proportion of drug-related deaths among the younger(15–49 years)male cohort,where the share of drug use disorders increased from 6.9%in 2000 to 11.8%in 2008.Conclusion:The deaths of despair crisis in the CEECA region is not a past event but an ongoing,evolving phenomenon.Its changing nature demands a shift in public health focus from solely historical drivers to new,generation-specific threats,particularly the rise of drug-related despair among youth.
基金supported by the National Natural Science Foundation of China(Grant No.82220108002 to F.C.and Grant No.82273737 to R.Z.)the U.S.National Institutes of Health(Grant Nos.CA209414,HL060710,and ES000002 to D.C.C.,Grant Nos.CA209414 and CA249096 to Y.L.)+1 种基金the Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD)supported by the Qing Lan Project of the Higher Education Institutions of Jiangsu Province and the Outstanding Young Level Academic Leadership Training Program of Nanjing Medical University.
文摘Interferon-related genes are involved in antiviral responses,inflammation,and immunity,which are closely related to sepsis-associated acute respiratory distress syndrome(ARDS).We analyzed 1972 participants with genotype data and 681 participants with gene expression data from the Molecular Epidemiology of ARDS(MEARDS),the Molecular Epidemiology of Sepsis in the ICU(MESSI),and the Molecular Diagnosis and Risk Stratification of Sepsis(MARS)cohorts in a three-step study focusing on sepsis-associated ARDS and sepsis-only controls.First,we identified and validated interferon-related genes associated with sepsis-associated ARDS risk using genetically regulated gene expression(GReX).Second,we examined the association of the confirmed gene(interferon regulatory factor 1,IRF1)with ARDS risk and survival and conducted a mediation analysis.Through discovery and validation,we found that the GReX of IRF1 was associated with ARDS risk(odds ratio[OR_(MEARDS)]=0.84,P=0.008;OR_(MESSI)=0.83,P=0.034).Furthermore,individual-level measured IRF1 expression was associated with reduced ARDS risk(OR=0.58,P=8.67×10^(-4)),and improved overall survival in ARDS patients(hazard ratio[HR_(28-day)]=0.49,P=0.009)and sepsis patients(HR_(28-day)=0.76,P=0.008).Mediation analysis revealed that IRF1 may enhance immune function by regulating the major histocompatibility complex,including HLA-F,which mediated more than 70%of protective effects of IRF1 on ARDS.The findings were validated by in vitro biological experiments including time-series infection dynamics,overexpression,knockout,and chromatin immunoprecipitation sequencing.Early prophylactic interventions to activate IRF1 in sepsis patients,thereby regulating HLA-F,may reduce the risk of ARDS and mortality,especially in severely ill patients.
基金Supported by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R896).
文摘Deep neural networks have achieved excellent classification results on several computer vision benchmarks.This has led to the popularity of machine learning as a service,where trained algorithms are hosted on the cloud and inference can be obtained on real-world data.In most applications,it is important to compress the vision data due to the enormous bandwidth and memory requirements.Video codecs exploit spatial and temporal correlations to achieve high compression ratios,but they are computationally expensive.This work computes the motion fields between consecutive frames to facilitate the efficient classification of videos.However,contrary to the normal practice of reconstructing the full-resolution frames through motion compensation,this work proposes to infer the class label from the block-based computed motion fields directly.Motion fields are a richer and more complex representation of motion vectors,where each motion vector carries the magnitude and direction information.This approach has two advantages:the cost of motion compensation and video decoding is avoided,and the dimensions of the input signal are highly reduced.This results in a shallower network for classification.The neural network can be trained using motion vectors in two ways:complex representations and magnitude-direction pairs.The proposed work trains a convolutional neural network on the direction and magnitude tensors of the motion fields.Our experimental results show 20×faster convergence during training,reduced overfitting,and accelerated inference on a hand gesture recognition dataset compared to full-resolution and downsampled frames.We validate the proposed methodology on the HGds dataset,achieving a testing accuracy of 99.21%,on the HMDB51 dataset,achieving 82.54%accuracy,and on the UCF101 dataset,achieving 97.13%accuracy,outperforming state-of-the-art methods in computational efficiency.
基金supports from the National Natural Science Foundation of China(No.52304148)the Youth Project of Shanxi Basic Research Program(No.202203021212262).
文摘This study investigates the performance enhancement of super-sulfated cement(SSC)derived from arsenic-containing bio-oxidation waste(BW)through the incorporation of carbonated recycled concrete fines(CRCF).The findings revealed that the addition of 5wt%CRCF yields optimal performance,with compressive strengths reaching approximately 1.83,12.59,and 42.81 MPa at 1,3,and 28 d,respectively.These values represented significant increases of 408.3%,10.0%,and 14.3%compared to the reference sample.The improvement was attributed to the synergistic effects of ultrafine CRCF particles acting as fillers and nucleation sites,as well as the high reactivity of silica gels,which promoted the formation of additional hydration gels.Microstructural analysis confirmed that CRCF addition refined pore structure,and enhanced the stiffness of C-S-H gels.Furthermore,CRCF served as a net CO_(2) sink,sequestering 0.268 kg CO_(2) per kilogram of CRCF and thereby reducing the carbon footprint of SSC.In addition,the feasibility of applying CRCF-modified SSC in cemented paste backfill(CPB)is highlighted,given the high cement-related carbon footprint of conventional CPB.When 5wt%CRCFmodified SSC was employed in CPB,its 3-d compressive strength attained over 70%of that of ordinary Portland cement(OPC),while the 28-d strength was comparable to that of OPC.The proposed binder thus provides a sustainable pathway for BW valorization,combining waste utilization,carbon sequestration,and improved engineering performance.
基金supported by a Project Grant(Grant No.PJT183705)an Early Career Investigator Prize(Grant No.ECP 184184)from the Canadian Institutes of Health Research+7 种基金a Prentice Institute Research Affiliate Fund Grant from the Prentice Institute for Global Population and Economy(Grant No.G00004116)a Te Herenga Waka Victoria University of Wellington Division of Science Health Engineering Architecture and Design Innovation Faculty Strategic Research Grant(Grant No.FSRG-SHEADI-10724)The Thailand Physical Activity Knowledge Development Centre(TPAK)/Thai Health Promotion Foundation provided funding for the cognitive interviews and pilot study in Thailand(Grant No.66-P1-0473)The University Pablo de Olavide provided a scholarship for 2 undergraduate students working on the project(codes PPI2207 and PPI2308)In the Czech Republicthe study was supported by Palacky University IGA(Grant No.IGA_FTK_2023_017)supported by the Division of Intramural Research at the National Institute on Minority Health and Health Disparities of the National Institutes of Healthsupported by the Key Project of the National Philosophy and Social Science Foundation of China(23&ZD197)。
文摘Background:Investigators from low-,middle-,and high-income countries representing 6 continents contributed to the development of the Global Adolescent and Child Physical Activity Questionnaire(GAC-PAQ).The GAC-PAQ is designed to assess physical activity(PA)across all key domains(i.e.,school,chores,work/volunteering,transport,free time,outdoor time).It aimed to address multiple gaps in global PA surveillance(e.g.,omission of important PA domains,insufficient cultural adaptation,underrepresentation of rural areas in questionnaire validation studies).The purpose of this study was to assess the content validity of the GAC-PAQ among PA experts,8-to 17-year-olds,and one of their parents/guardians,and to discuss changes made to the questionnaire based on participants'feedback.Methods:Sixty-two experts in PA measurement and/or surveillance from 24 countries completed an online survey that included both closed-and open-ended questions about the content validity of the GAC-PAQ.The proportion of experts who agreed or strongly agreed with the items was calculated.Child-parent/guardian dyads from 15 countries(n=250;10-40 per country)participated in a structured cognitive interview to assess the clarity of the questions and response options,and they were encouraged to provide suggestions to improve clarity and facilitate completion of the questionnaire.Participating countries are:Aotearoa New Zealand,Brazil,Canada,China,Colombia,Czech Republic,India,Malawi,Mexico,Nepal,Nigeria,Spain,Sweden,Thailand,and the United Arab Emirates.Interviews were conducted in 13 different languages and structured by PA domain.Generic images were included to help participants in answering questions about PA intensity.Results:Expert agreement with the items for each domain exceeded 75%,and their qualitative feedback was used to revise the questionnaire before cognitive interviews.In general,participants found the questionnaire to be comprehensive.Adolescents(12-17 years)found it easier than children(8-11 years)to answer the questions.Several children struggled to answer questions about the duration and intensity of activities and/or concepts related to travel modes,active trips,and organized activities.Many parents/guardians were unsure about the frequency,duration,and intensity of their children's or adolescents'PA at school and/or recommended using more culturally relevant and appropriate images.Some participants misunderstood the concept of activities that“make you stronger”(intended to assess resistance activities)and/or struggled to differentiate between work,volunteering,and chores.Conclusion:Participants'feedback was used to develop a revised,simplified,and culturally adapted GAC-PAQ,which will be pilot-tested in all15 countries in an App that will include country-specific images and narration in local languages.Further research is needed to assess the reliability and validity of the revised GAC-PAQ.
基金supported by the National Natural Science Foundation of China Grant(No.61972133)Project of Leading Talents in Science and Technology Innovation for Thousands of People Plan in Henan Province Grant(No.204200510021)the Key Research and Development Plan Special Project of Henan Province Grant(No.241111211400).
文摘At inference time,deep neural networks are susceptible to backdoor attacks,which can produce attackercontrolled outputs when inputs contain carefully crafted triggers.Existing defense methods often focus on specific attack types or incur high costs,such as data cleaning or model fine-tuning.In contrast,we argue that it is possible to achieve effective and generalizable defense without removing triggers or incurring high model-cleaning costs.Fromthe attacker’s perspective and based on characteristics of vulnerable neuron activation anomalies,we propose an Adaptive Feature Injection(AFI)method for black-box backdoor detection.AFI employs a pre-trained image encoder to extract multi-level deep features and constructs a dynamic weight fusionmechanism for precise identification and interception of poisoned samples.Specifically,we select the control samples with the largest feature differences fromthe clean dataset via feature-space analysis,and generate blended sample pairs with the test sample using dynamic linear interpolation.The detection statistic is computed by measuring the divergence G(x)in model output responses.We systematically evaluate the effectiveness of AFI against representative backdoor attacks,including BadNets,Blend,WaNet,and IAB,on three benchmark datasets:MNIST,CIFAR-10,and ImageNet.Experimental results show that AFI can effectively detect poisoned samples,achieving average detection rates of 95.20%,94.15%,and 86.49%on these datasets,respectively.Compared with existing methods,AFI demonstrates strong cross-domain generalization ability and robustness to unknown attacks.
基金supported by the Advanced Materials-National Science and Technology Major Project(Grant No.2025ZD0618401)the National Natural Science Foundation of China(Grant No.12504285)+1 种基金the Natural Science Foundation of Jiangsu Province(Grant No.BK20250472)NFSG grant from BITS-Pilani,Dubai campus。
文摘The rapid advancement of machine learning based tight-binding Hamiltonian(MLTB)methods has opened new avenues for efficient and accurate electronic structure simulations,particularly in large-scale systems and long-time scenarios.This review begins with a concise overview of traditional tight-binding(TB)models,including both(semi-)empirical and first-principles approaches,establishing the foundation for understanding MLTB developments.We then present a systematic classification of existing MLTB methodologies,grouped into two major categories:direct prediction of TB Hamiltonian elements and inference of empirical parameters.A comparative analysis with other ML-based electronic structure models is also provided,highlighting the advancement of MLTB approaches.Finally,we explore the emerging MLTB application ecosystem,highlighting how the integration of MLTB models with a diverse suite of post-processing tools from linear-scaling solvers to quantum transport frameworks and molecular dynamics interfaces is essential for tackling complex scientific problems across different domains.The continued advancement of this integrated paradigm promises to accelerate materials discovery and open new frontiers in the predictive simulation of complex quantum phenomena.
