Browse Articles
Discover research articles across all indexed journals
An efficient hybrid reliability analysis method with application to the harmonic drive
The harmonic drive is a new type of high-efficiency gear transmission system, widely used in precision machinery, robotics, and aerospace due to its compact structure and high transmission accuracy, and the multiple uncertain parameters in it will seriously affect its transmission performance. In this paper, the limit state functions for multiple failure modes of harmonic drive are established based on the stress-strength interference theory. Considering the hybrid uncertainties in the harmonic drive, the modified chaos control method and multiplicative dimension reduction method are coupled to solve the probabilistic and interval reliabilities continuously. The proposed model is simple and easy to implement, and the failure probabilities of different failure modes of harmonic reducers can be accurately and quickly evaluated. Furthermore, the differences in the influence of strength on different failure modes are demonstrated by the results of hybrid uncertainty analysis.
NMR based clinical metabolomics revealed altered lipid and sugar metabolism in systemic sclerosis with implications for diagnosis and therapeutic strategies
Predicting art university students’ entrepreneurial intention: A hybrid SEM–ANN approach
In recent years, academics and policymakers have increasingly focused on entrepreneurial behavior among university students. While existing studies have explored the entrepreneurial intention (EI) of students from various academic disciplines, few have specifically examined the EI of art university students. Based on the Diffusion of Innovations Theory (DOI) and the Theory of Planned Behavior (TPB), this study explores the factors influencing art university students’ EI and assesses each factor’s relative importance. This study employed a structural questionnaire to survey 273 students from three universities in Liaoning Province, China, measuring eight constructs: relative advantage (RA), observability (OB), compatibility (CO), entrepreneurial motivation (EM), entrepreneurial attitude (EA), subjective norms (SN), perceived behavioral control (PBC), and EI. Data analysis was conducted using Structural Equation Modeling (SEM) and Artificial Neural Networks (ANN). The results show that, among the direct significant predictors of EI, PBC has the strongest influence, followed by EA, SN, and EM. Additionally, all predictive constructs accounted for 60% of the variance in the EI of art university students. The ANN analysis revealed the following normalized importance ranking of all predictive constructs: PBC (100%), EA (70.8%), SN (57.6%), RA (43.1%), and EM (31.2%). This study not only fills the gap in research on the EI of art university students but also provides valuable insights for developing targeted strategies to foster entrepreneurship among this group.
Semi-analytical solution of cohesive zone model for cement-based materials
Predicting customer loyalty in omnichannel retailing using purchase behavior, socio-cultural factors, and learning techniques
In the competitive retail omnichannel market, customer loyalty is essential for maintaining market share and reducing the cost of acquiring new customers. Previous research has primarily focused on factors influencing customer loyalty, often in isolation, but this study goes beyond traditional approaches. The aim of this research is to fill significant gaps in current studies by integrating a more comprehensive set of variables that reflect the complex and dynamic nature of customer loyalty in a flexible omnichannel environment. The main innovation of this study lies in the use of new and comprehensive omnichannel data, which includes sales data across various platforms, socio-economic conditions, shopping cart behaviors, and customer sentiments. The proposed model utilizes a hybrid approach, incorporating BERT for sentiment analysis, reinforcement learning for behavior analysis, and fine-tuning for improved predictions. Additionally, graph-based models (GCN) and adaptive learning are employed to analyze trends and predict loyalty at both individual and neighborhood levels. This research provides an intelligent analytical framework for predicting customer loyalty in omnichannel retail environments, enhancing Customer Relationship Management (CRM) subsystems within Enterprise Information Systems (EIS). By optimizing decisions in areas such as pricing, inventory management, and personalized advertising, this study ultimately leads to improved customer retention and increased market competitiveness.
Biological activities and application of Rosmarinus officinalis extract to improve the preservation and microbial qualities of some local meat products
Abstract Because medicinal plants contain bioactive phenolic compounds with antibacterial, antioxidant, and other properties, they have been used in many parts of nutrition and healthcare. The atmosphere in which local meat products and vendors operate is still subpar. Every individual is impacted, either directly or indirectly, when food is contaminated by harmful germs and spoilage. In this study, the microbiological purity, preservation, and shelf life of beef meat were assessed. The study analyzed the bacterial count in beef burger and luncheon samples, with gram negative bacteria having the highest contamination rate (58%). Four selected bacteria isolates were identified based on their morphological and cultural characteristics, confirmed by VITEK 2 system analysis and were identified as bacterial isolates from the Bacillaceae, Enterobacteriaceae, Pseudomonadaceae, and Staphylococcaceae families. The study found nine active compounds in rosemary leaves extract, including ferruginol, camphor, cineole, verbenone, borneol, αcaryophyllene, terpinen-4-ol, 2-methyl-4-vinylphenol, and eugenol, which act as antioxidants and antibacterials. The aqueous rosemary leaves extract (ARLE) showed efficacy in antibacterial activity against various bacteria, with clear inhibition zones (CIZ) and mean growth inhibition (MGI) values of 28.1 mm and 37%, respectively. The addition of ARLE to meat beef significantly reduced counts of pathogenic bacteria during storage. The preservation effectiveness of ARLE against artificially inoculated Staphylococcus aureus and Pseudomonas aeruginosa in meat beef exhibited complete bactericidal effect, with no recovery of the pathogen. The study assessed the impact of ARLE on meat beef sensory quality, finding that at a concentration of 15 mg/g, ARLE did not affect overall acceptability but enhanced the meat’s sensory properties. When used as a natural antioxidant agent, the ARLE’s antioxidant activity through DPPH scavenging proved to be successful in food preservation. Its radical-scavenging activity increased with concentration, with high DPPH radical-scavenging activity at 2560 μg/mL and varying percentages at different concentrations. ARLE also showed cytotoxicity against colon carcinoma cells (HCT-116), breast carcinoma cells (MCF-7), and human hepatocellular carcinoma cells (HepG-2) for 24 h. Applying ARLE significantly extended meat shelf-life by reducing microbial counts and inhibiting pathogens. To strengthen this conclusion, referencing specific safety and spoilage thresholds, such as TVC below 10^6 CFU/g, would provide a clearer, quantitative assessment of preservation. Demonstrating that treated samples remained below these limits longer than controls would offer concrete evidence of practical shelf-life extension.
Malnutrition among under-five children in amhara and oromia regions, Ethiopia: Continuous time markov multi-state modeling
Background Ethiopia faces a high burden of undernutrition prevalence, ranking among the 15 worst-affected nations globally. So, this study aimed to find out how often and how long under-five children (U5C) in Ethiopia’s Amhara and Oromia regions move between different states of Composite Index Anthropometry Failure (CIAF) as well as what factors influence these changes. Methods The data used for this study was extracted from the International Food Policy Research Institute. The institute conducted a follow-up survey in three consecutive rounds: February 8, 2018–April 25, 2018; July 25, 2019–October 23, 2019; and February 8, 2021–April 25, 2021, respectively. The inclusion criteria were households that had children between the ages of 0 and 35 months, were participants in the safety net program, and had the mother or primary female caregiver during the baseline survey. A total of 3,044 households having children with at least two complete anthropometric measurements were included. A continuous-time multi-state Markov model was used to estimate transitions and their probability between CIAF’s states. Results Nourished children had a 71% probability of becoming undernourished. Time taken to recover from undernourished state for U5C was 41 months on average. 75% of the U5C’s life is spent in the undernourished state. Girls had a 1.824 times higher likelihood of recovering from an overnourished state and were less likely to transit from a nourished state to an undernourished state compared to boys (HR: 0.8013). Children older than two years were more likely to recover from undernourished and overnourished states respectively (HR: 1.013 & 1.036), to a nourished state and less likely to transit back to the malnutrition states (HR: 0.9693 & 0.9662). Children of educated mothers and residents in the Oromia region had lower risk of transition from a healthy to an undernourished state respectively (HR: 0.8171& 0.8074). Conclusion Undernutrition will affect most U5Cs. Children whose mothers had no education and live in the Amhara region are more susceptible to undernutrition. The Ministry of Health and other relevant stakeholders should develop a practical intervention to enhance adult and maternal education programs.
Correction: Comparing the mechanical energetics of walking among individuals with unilateral transfemoral limb loss using socket and osseointegrated prosthetic interfaces
Estimating the size of hard to sample populations: A comprehensive study on female sex workers and sexually exploited minors in Rwanda using privatized network sampling in 2023
Introduction Female sex workers (FSW) are at increased risk of HIV and other STI. In addition, the burden of HIV infection among this group is much higher when compared to adult females in the general population. Estimating the number of FSW helps HIV/STI prevention through program design, planning, and implementation. The aims of this study are to provide the most up to date national population size estimates (PSE) and geographical distribution of female sex workers and sexually exploited minors in Rwanda. Having population size estimates of the HIV-mostly affected population, FSW in this case provides the basis for determining the denominators to assess HIV program performance towards national and global targets of controlling the HIV epidemic among the FSW population. Methods Data were collected from May 8th to June 24th, 2023, across 10 study sites countywide. Privatized network sampling (PNS) was used, which is a population size estimation method that uses the network information collected within a bio-behavioral survey (BBS) that used respondent-driven sampling (RDS). To estimate the FSW and sexually exploited minors’ population size, three PNS estimators were used: Cross-Sample, Cross-Alter, and Cross-Network. Results The national-level FSW population size was estimated at 98,587 (95% CI: 82,978–114,196), corresponding to 2.3% of the total adult female population aged 15 years and above in Rwanda. We estimated that in the City of Kigali, 5.3%, in the West Province, 2.2%, in the East and South province, 1.7% each, and in the North province 1.6% of adult female population 15 years of age and older who were FSW. Conclusion This was the first time that PNS was implemented as a PSE method in Rwanda, adding to the emerging tools that we have in the hard-to-reach PSE field. The PSE provides fundamental information to design, plan, and implement programs for FSW at the provincial level in Rwanda. Furthermore, these estimates will help to generate positive policy changes and to advocate for resources that will help in the effort to achieve a sustained HIV epidemic control in the country.
HLA-B*58:01 genotyping prevalence and the association with allopurinol-induced severe cutaneous adverse reactions: a living systematic review and meta-analysis
Tick phenology, tick-host associations, and tick-borne pathogen surveillance in a recreational forest of East Texas, USA
Management of tick-borne disease necessitates an understanding of tick phenology, tick-host associations, and pathogen dynamics. In a recreational hotspot outside of one of the largest cities in the United States, we conducted a year of monthly standardized tick drag sampling and wildlife trapping in Sam Houston National Forest, a high use recreation site near Houston in east Texas, US. By sampling 150 wildlife hosts of 18 species, including rodents, meso-mammals, deer, reptiles, and amphibians, we collected 87 blood samples, 90 ear biopsies, and 861 ticks representing four species (Amblyomma americanum, Dermacentor variabilis, Ixodes scapularis and Ixodes texanus). Drag sampling yielded 1,651 questing ticks of three species: A. americanum (921), D. variabilis (10), and I. scapularis (720). Off-host larval A. americanum abundance peaked in July, followed by peak infestations of wildlife, predominantly raccoons, in August. Off-host I. scapularis larvae abundance peaked in spring (March-May), while very few were removed from hosts and only a single I. scapularis nymph was found throughout the study via dragging in June. In contrast, both off-host and on-host adult I. scapularis occurred most frequently in the winter. Overall, tick infections included 25.3% (183/725) with Rickettsia buchneri, 15.5% (112/725) Rickettsia amblyommatis, 8.0% (58/725) Rickettsia tillamookensis, 0.8% (6/725) Rickettsia spp., and a single tick with a hard tick relapsing fever Borrelia spp.; no tick tested positive for Borrelia burgdorferi. Characterizing tick phenology, tick-host associations, and tick-borne bacteria fills important knowledge gaps for the risk of tick-borne diseases in pine-dominated forests of this region.
Y665F variant of mouse Stat5b protects against acute kidney injury through transcriptomic shifts in renal gene expression
Abstract The impact of single nucleotide polymorphisms (SNP) on physiology is often underestimated. One amino acid change can result in a variety of phenotypes apparent only in response to disease or injury. Even known pathogenic SNPs have widespread effects that are currently unaccounted for. In this study, we investigated the impact of the known activating and pathogenic Stat5b Y665F mutation in a renal injury context in mice carrying this variant. Using ischemia–reperfusion model of acute kidney injury, immunohistochemistry, RNA-seq and ChIP-seq, we establish the protective role of STAT5b activation in renal epithelium and showcase the shifts in transcriptomic landscape in a tissue not associated with the usual human phenotype of the STAT5B Y665F mutation. Our data indicate new links between the JAK/STAT pathway and known kidney injury markers, contribute to the understanding of the sexual dimorphism of renal disease, and provide new potential targets for JAK inhibitor- and amino acid transport modulation-based therapies.
Multi-task deep learning for predicting metabolic syndrome from retinal fundus images in a Japanese health checkup dataset
Background Retinal fundus images provide a noninvasive window into systemic health, offering opportunities for early detection of metabolic disorders such as metabolic syndrome (METS). Objective This study aimed to develop a deep learning model to predict METS from fundus images obtained during routine health checkups, leveraging a multi-task learning approach. Methods We retrospectively analyzed 5,000 fundus images from Japanese health checkup participants. Convolutional neural network (CNN) models were trained to classify METS status, incorporating fundus-specific data augmentation strategies and auxiliary regression tasks targeting clinical parameters such as abdominal circumference (AC). Model performance was evaluated using validation accuracy, test accuracy, and the area under the receiver operating characteristic curve (AUC). Results Models employing fundus-specific augmentation demonstrated more stable convergence and superior validation accuracy compared to general-purpose augmentation. Incorporating AC as an auxiliary task further enhanced performance across architectures. The final ensemble model with test-time augmentation achieved a test accuracy of 0.696 and an AUC of 0.73178. Conclusion Combining multi-task learning, fundus-specific data augmentation, and ensemble prediction substantially improves deep learning-based METS classification from fundus images. This approach may offer a practical, noninvasive screening tool for metabolic syndrome in general health checkup settings.
Mapping Alzheimer’s disease pathology using free water through integrated analysis of plasma biomarkers, microstructural DTI metrics, and macrostructural MRI measures
Effectiveness of a multilevel intervention to improve mental health of hospital workers: The SEEGEN multicenter cluster randomized controlled trial
Introduction Hospital workers are at high risk for stress-related mental health issues and are considered a vulnerable workforce in most Western countries. Although multilevel interventions that address individual and organizational factors show promise, there is limited robust evidence of their effectiveness in hospital settings. This study evaluated the SEEGEN trial, a cluster-randomized controlled trial conducted in the German healthcare sector, to assess the effectiveness of a structured multilevel intervention designed to reduce psychosocial stress and to promote mental well-being among hospital employees. The intervention included five modules that targeted different hierarchical levels, sources of interpersonal and structural stress, and potentially vulnerable life stages. These modules were: (i) top management training, (ii) dilemma management – coping by taking responsibility, (iii) promoting stress-preventive relational leadership competence, (iv) reconciling work and family life, and (v) staying healthy at work. Methods The study was conducted at three clinical centers in Germany and included 18 clusters with a total of N = 415 participants. The clusters were randomly assigned to either an intervention or a wait-list control group. The primary outcome was psychological strain (Irritation Scale; IRR), and the secondary outcomes were mental well-being (WHO-5) and perceived psychosocial safety climate, (PSC-12). Intervention effects were estimated using a two-level linear analysis of covariance. Changes from baseline to the 11-month follow-up were analyzed. Results The intervention had no statistically significant effect on the primary or secondary outcomes. Conclusions The lack of significant effects may be attributed to low participation rates, an insufficient intervention dosage, and contextual factors, such as the SARS-CoV-2 pandemic and staffing shortages in the participating hospitals. Although the intervention cannot currently be recommended for widespread implementation, the study provides valuable insights into developing, delivering, and overcoming the challenges of multilevel workplace interventions in healthcare settings.
The mediating effect of social adaptation and the moderating effect of prosocial behavior on the relationship between physical activity and psychological capital
Herbicide dose-response thresholds in sands to assess the risk of non-target damage to winter grain crops
Herbicide residues in soil from previous crops or from pre-emergent treatments can have unintended toxicity on the next crop. Despite this there is limited published information on toxicity thresholds for many crops or herbicides. This study aimed to quantify shoot and root responses of six common winter grains crops (canola, chickpea, fieldpea, lentil, lupin and wheat) to increasing concentration of four common herbicides (clopyralid, pyroxasulfone, propyzamide and trifluralin) in soil. Lentil emergence was highly sensitive to clopyralid (29 μg kg-1 for a 50% reduction, ED50) while wheat emergence was sensitive to propyzamide and trifluralin, with complete inhibition at 100 μg kg-1 and 375 μg kg-1, respectively. Shoot and root parameters of the legumes, except lupin, were significantly reduced by clopyralid, with ED50 values ranging between 3−27 μg kg-1. Canola was sensitive to pyroxasulfone, with shoot and root biomass ED50 at 21 and 8 μg kg-1, respectively. Pyroxasulfone also severely reduced root length of all tested crops (ED50 values 6−53 μg kg-1). Root and shoot growth in wheat was most susceptible to propyzamide followed by trifluralin. This study found that one or more herbicides had the potential to cause significant phytotoxic effects in all crops at concentrations below recommended application rates and below those detected in a recent field survey of pre-sowing herbicide residues in field soils around Australia. These results suggest the risk of early crop damage residual herbicides in very light-textured soils. More effort is now required to determine potential effects on different soil types and crop yields, to enable better spatial and economic risk assessment.
Understanding and predicting flossing behavior in a rural Appalachian community using the multi-theory model
A hybrid approach for forecasting peak expiratory flow rate in asthma patients using combined linear regression and random forest model
Asthma is a frequent and long-lasting disorder associated with airway inflammation. The disease severity may lead to serious health concerns and even mortality. In this work, we propose a novel hybrid approach using machine learning models and similarity measurement technique with the aim of precise peak expiratory flow rate (PEFR) estimation for asthma trigger assessment. The random forest model was first utilized to classify the PEFR percentile zones on unseen data. Then, two linear regression models following thresholds of <50% and >=50% were hypothesized and trained to achieve better outcomes than a single standalone model. Hence, the input is diverted to the relevant model for prediction based on classification results. Furthermore, a string-matching technique has been proposed to obtain reference outcomes in addition to yesterday’s PEFR. Finally, a supplementary linear regression model is used to make predictions based on input of two prediction values and one PEFR value from the previous day. The proposed model is evaluated on a dataset of 25 patients, each with 2 to 3 months of recordings, on average. The findings showed reduced mean and random absolute error of 27.064 L/min and 1.34%, respectively, using the suggested model, compared to 79.794 L/min and 4.42% error rates by the standalone linear regression model on five-fold cross-validation. The outcome indicates that the proposed hybrid algorithm accurately predicts asthma-trigger events.