Browse Articles

Multistage feature selection and stacked generalization model for cancer detection

Environmental management accounting affects bank performance with mediators

This study aims to estimate the effect of environmental management accounting (EMA) on Vietnamese bank performance under the mediating role of environmental costs. The research employs qualitative methods, such as expert interviews and surveys, alongside quantitative methods, such as Partial least squares Structural Equation Modeling (PLS-SEM). The findings revealed a positive correlation between EMA and knowledge management, green innovation, and environmental costs. Additionally, knowledge management and green innovation significantly positively influence environmental costs. Significantly, the study emphasizes the relationship between EMA and a bank’s performance, mediated by environmental costs. Recognizing the significance of environmental costs in the total cost structure, this study highlights their potential emergence in the provision of financial services. This study underscores the role of environmental accounting, which integrates the financial and management accounting aspects, in providing information on these costs.

MicroRNA expression in peripheral blood of children with asthma and the role of miR-31-5p in childhood asthma progression

The group-housed pigs attacking and daily behaviors detection and tracking based on improved YOLOv5s and DeepSORT

Automatic detection and tracking of pig behaviors through video surveillance remain challenges due to farm demanding conditions, e.g., illumination conditions and occlusion of one pig from another. The main goal of this study is to develop a deep learning method based on the improved YOLOv5s and DeepSORT to detect and track the behaviors of pigs, which has the advantages of stability and high accuracy. Firstly, YOLOv5s with the attention mechanism is used for pig detection and behavior recognition. To deal with the missed detection and false detection due to occlusion and overlapping between pigs and pigs, the improved YOLOv5s adopts the Shape-IoU to optimize the bounding box regression loss function, which improves the robustness of the model. Then, the improved DeepSORT model is proposed to track each pig behaviors including eat, stand, lie and attack four behavior types. Finally, we conduct a comparison test under different lighting and density conditions for pig detection and behavior tracking on special dataset. Experimental results show that the mAP@0.5% of improved YOLOv5s algorithm increases from 92.7% to 99.3%, which means 6.6% accuracy improvement compared with the YOLOv5s model. In terms of tracking, the values of MOTA and MOTP in all test videos are 94.5% and 94.9% respectively. These experiments demonstrate that the improved YOLOv5s and DeepSORT achieves high accuracy for both pig detection and behavior tracking. The proposed approach provides scalable technical support for contactless automatic pig monitoring.

A multimodal multitask deep learning model for predicting stroke lesion and functional outcomes using 4D CTP imaging and clinical metadata

HybridoNet-Adapt: A domain-adapted framework for accurate lithium-ion battery RUL prediction

Accurate prediction of the Remaining Useful Life (RUL) of lithium-ion batteries is critical for safe, reliable Battery Health Management in diverse operating conditions. Existing RUL models often fail to generalize when test data diverge from the training distribution. To address this, we introduce HybridoNet-Adapt, a domain-adaptive RUL prediction framework that explicitly bridges the gap between labeled source and unlabeled target domains. During training, we minimize the Maximum Mean Discrepancy (MMD) between feature distributions to learn domain-invariant representations. Simultaneously, we employ two parallel predictors—one tailored to the source domain and one to the target domain—and balance their outputs via two learnable trade-off parameters, enabling the model to dynamically weight domain-specific insights. Our architecture couples this adaptation strategy with LSTM, multi-head attention, and Neural ODE blocks for deep temporal feature extraction, but its core novelty lies in the MMD-based alignment and hybrid prediction mechanism. On two large, publicly available battery datasets, HybridoNet-Adapt consistently outperforms non-adaptive baselines (Structural Pruning, Multi-Time Scale Feature Extraction Hybrid model, XGBoost, Elastic Net), archiving an RMSE reduction of up to 152 cycles under domain shifts. These results demonstrate that incorporating domain adaptation into RUL modeling substantially enhances robustness and real-world applicability.

Enhanced photocatalytic degradation of organic pollutants using a TiO2–clay nanocomposite in a rotary photoreactor with experimental and theoretical insights

Insights from the care home staff on the use of observational risk assessment of contractures: Longitudinal evaluation (ORACLE): A qualitative study

Background Contractures are a common but preventable consequence of immobility and inactivity among residents living in care homes. There is a rising prevalence and subsequent impact of contractures on care home residents, leading to the development of a risk assessment tool for contractures, Observational Risk Assessment for Contractures: Longitudinal Evaluation (ORACLE). This qualitative study aims to explore the experience of care staff regarding the usability, acceptability, and practical implementation of ORACLE. Methods A qualitative study using a partly deductive and pragmatic approach was conducted through semi-structured interviews with care home staff in England. The care staff members were selected via purposive sampling and were interviewed either through videoconferencing or in person in a private room at care homes. The interviews were recorded and transcribed verbatim. The data collected was coded using NVivo and synthesised using thematic analysis. Results Ten care staff members were interviewed from five care homes (four senior staff members and six healthcare assistants). Three overarching themes were identified: 1) usability of ORACLE, 2) acceptability of ORACLE and 3) contextual factors that can potentially influence the practical implementation of ORACLE in a care home setting. Respondents found the tool to be user-friendly and well-integrated within existing care routines. The study also identified factors relating to care home processes, the people involved, the training environment, and the policy context that tend to support or inhibit the effective implementation of ORACLE. Conclusion The study offers preliminary insights into the usability and acceptability of ORACLE and its application in a care home setting.

Automated forest land division using deep learning and drone imagery

This paper proposes an automated solution for tree enumeration in areas designated for forest land division using drone image processing. Traditional tree counting methods are time-consuming and error-prone. Our approach leverages drone imagery and advanced computer vision algorithms. The solution demonstrates the potential to accurately detect tree crowns, facilitating informed decision-making in forest land division projects, promoting sustainability and efficient resource management.

Seawater-resistant emulsified epoxy resin for effective sand control in unconsolidated sandstone oil reservoir

Sand production in oil wells is recognized as a persistent challenge during oilfield development, adversely affecting well productivity and operational stability. Chemical sand control methods, particularly resin-based sand consolidation, are considered a promising solution due to their operational simplicity and effectiveness. However, conventional emulsified resins are known to be highly sensitive to high-salinity environments, which can lead to emulsion destabilization and reduced consolidation strength. To address this limitation, a novel emulsified epoxy resin system was developed in this study using a nonionic emulsifying curing agent—fatty amine poly(epoxy ethyl ether)—by which salinity tolerance is significantly enhanced, supporting dilution water salinity up to 3.8 × 10⁴ mg/L. Through single-factor experiments, an optimal formulation was identified as 16% epoxy resin, 24% emulsified curing agent, 1% coupling agent, and 5.6% stabilizer. The molecular structure of the emulsified resin and the stability of the cured matrix were thoroughly characterized. The effects of curing temperature, time, sand particle size, and stabilizer dosage on compressive strength and permeability were systematically evaluated. It was demonstrated that after being cured at 80 °C for 12 hours, the consolidated cores achieved a compressive strength exceeding 3 MPa with permeability retention above 75%. Furthermore, the consolidated cores were shown to exhibit excellent long-term stability, maintaining their mechanical and flow properties after 30-day immersion in kerosene, 10% HCl, and formation water. This study bridges a critical research gap in high-salinity applications of water-based resin emulsions and provides a robust technical solution for sand control in challenging reservoir environments.

MALDI-TOF MS for malaria vector surveillance: A cost-comparison analysis using a decision-tree approach

Background The use of MALDI-TOF MS for mosquito identification and surveillance is routinely used in developed countries as an affordable alternative to molecular methods. However, in low- and middle-income countries (LMIC) where mosquito-borne diseases carry the greatest burden, the method is not commonly employed. Using the Kenyan national malaria program (NMCP) as a case study, we compared the costs of current methods used for malaria vector surveillance to those that would be incurred if MALDI-TOF MS were used instead. Methods A deterministic decision tree analytic model was developed to systematically calculate the costs associated with materials and labour, and time-to-results for two workflows, i.e., current molecular methods versus MALDI-TOF MS. The analysis assumed an annual sample size of 15,000 mosquitoes (representing the average number of mosquitoes analysed annually by the Kenyan NMCP) processed at a local laboratory in Kenya. Findings We estimate that if the Kenyan national entomological surveillance program shifted sample processing completely to MALDI-TOF MS, it would result in 74.48% net time saving, up to 84% on material costs and 77% on labour costs, resulting in an overall direct cost savings of 83%. Interpretation Adoption of MALDI-TOF MS for malaria vector surveillance can result in substantial time and cost savings. The ease of performance, the rapid turn-around time, and the modest cost per sample may bring a paradigm shift in routine entomological surveillance in Africa.

Estimated glucose disposal rate and risk of metabolic syndrome: A population-based study

Aims This study aims to explore the association between the estimated glucose disposal rate (eGDR) and the risk of metabolic syndrome (MetS), with a focus on the mediating role of BMI. Methods Data for this study came from the 2011 and 2015 China Health and Retirement Longitudinal Study. We used multivariable logistic regression and restricted cubic splines to assess the relationship between eGDR and MetS, with subgroup and interaction analyses to identify moderating factors. The diagnostic ability of eGDR for MetS was evaluated via receiver operating characteristic curves. Results In total, 3,229 participants were included, with 745 (23.07%) diagnosed with MetS. In the fully adjusted model, each interquartile range increase in eGDR was associated with a 58% reduced MetS risk (odds ratio = 0.42, 95% confidence interval: 0.36–0.49, P < 0.001). A significant nonlinear dose–response relationship between eGDR and MetS risk was observed (P < 0.001, both overall and nonlinear).Spline regression analysis revealed that the protective effect of eGDR was significant up to 11.88 mg/kg/min (standard error = 0.17). Subgroup analysis revealed significant interaction effects of marital status and residential area on the eGDR–MetS relationship (P < 0.05), while BMI mediated 29% of eGDR’s total effect on MetS. Receiver operating characteristic analysis showed eGDR’s good predictive performance for MetS, with an area under the curve of 0.71 (95% confidence interval: 0.69–0.73). Conclusion Higher eGDR levels were linked to a significantly lower MetS risk, with approximately 29% of this association mediated by BMI, suggesting that individuals with low eGDR may benefit from closer monitoring for MetS development.

Structure and function of Full-length Tau

Tau protein, encoded by the MAPT gene, is a microtubule-associated protein involved in the regulation of microtubule stability in neurons, contributing to cell shape maintenance and intracellular transport, among other functions. Tau is not found as a unique isoform; instead, different Tau isoforms of varying sizes are present in the brain, but a Full-length Tau isoform (Full Tau) containing all 16 exons has never been previously identified. This study has explored the structure and function of the Full Tau isoform, which includes all exons of the MAPT gene. To achieve this, we expressed the Full Tau isoform in bacteria, alongside the Tau 4R2N isoform as a control, and tested its microtubule-binding capacity, self-aggregation propensity, and effects on cultured cells regarding cell proliferation and cell death. Our results indicated several differences between the Full Tau and Tau 4R2N isoforms, suggesting distinct roles in cellular dynamics. To explain these differences, we suggest the role of exon 8, which is present in the Full Tau isoform but absent in Tau 4R2N.

Spatiotemporal evolution and influencing factors of agricultural carbon emissions in China

Clarifying the spatiotemporal characteristics of agricultural carbon emissions and influencing factors in China is crucial. A system for measuring agricultural carbon emissions was established, thus evaluating the level of carbon emissions in China and its provinces. Moreover, the dynamic evolution of agricultural carbon emissions in China and the regions on both sides of the Hu Line was analyzed, then investigated factors affecting agricultural carbon emissions by the LMDI model. The results indicate that the total amount and intensity of agricultural carbon emissions showed an upward and then a downward trend in China from 2001 to 2021. The peaks were 330.72 million tons and 1.98 tons\ha, respectively. Agricultural carbon intensity in provinces was mostly Low-Low Cluster and the range of High-High Cluster has decreased. Inter-provincial disparities in agricultural carbon emissions were also gradually narrowing. These show that the effect of agricultural carbon emissions reduction was obvious in China. It is important to note that carbon emissions from energy consumption in agriculture and agricultural material inputs were substantial, accounting for about 95% of the total. Agricultural carbon emissions were restricted by the agricultural production efficiency, changes in industrial structure, rural population size, and agricultural industrial structure, but were promoted by the level of economy and urbanization. Therefore, we recommend enhancing inter-provincial synergistic collaboration to create agricultural carbon emissions reduction pathways with unique features. It is also essential to maximize agricultural production efficiency and grasp the direction of green and low-carbon. We also suggest that the Chinese government should accelerate the in-depth adjustment and transformation and upgrading of the industrial structure, thereby reducing agricultural carbon emissions at source.

Geographic variations and trends in percutaneous intervention for patients with and without acute myocardial infarction: A Japanese nationwide registry study

Background A previous study demonstrated that the per capita volume of percutaneous coronary intervention (PCI) for acute myocardial infarction (AMI) was relatively uniform across the 47 prefectures in Japan, while elective PCIs for stable coronary artery disease showed wide regional variation. However, contemporary data remain limited. Methods The Japanese PCI (J-PCI) is a nationwide prospective registry covering most of the procedures performed within the country. PCI procedures in 2019 and 2023 were included and divided according to the indications: AMI versus non-AMI. The patterns of PCI procedures performed for AMI and non-AMI across all prefectures in Japan were evaluated. The associations of the non-AMI/AMI ratio with population, area, and the number of PCI-capable centers per prefecture were also assessed. Results A total of 494,746 PCI procedures were analyzed. The ratios between the highest and lowest prefectures were 4.0-fold in non-AMI and 1.9-fold in AMI in 2019 and 4.2-fold in non-AMI and 2.0-fold in AMI in 2023. The non-AMI/AMI ratio was positively correlated with the ratios of the number of PCI-capable centers to population and area per prefecture. Conclusions Geographic disparity was observed in the relative volume of PCI performed for AMI compared to non-AMI across Japan, potentially reflecting variations in the density of PCI-capable centers relative to the area and population of each prefecture.

The effect of socioeconomic status on academic achievement: A big data study across countries and time with integrative data analysis

The strong relationship between socioeconomic status (SES) and academic achievement has been the focus of much research over the last fifty years. However, there is a very limited number of studies that examine the SES of individual students and the SES of their peers in a disaggregated manner. This study examines the relationship between SES and academic achievement at both the student and school levels, using data from 54 countries from six cycles of the Trends in International Mathematics and Science Study (TIMSS) assessment over the past two decades. Integrative Data Analysis (IDA) is used to synthesize data across cycles and countries, and Hierarchical Linear Modeling (HLM) is used to account for the nested nature of students within schools. The results show that both individual and school SES have a significant positive effect on academic achievement, with school SES having a stronger effect. Moreover, most of the variance in the effects of individual and school SES is due to differences between countries. Income inequality reduces the effect of individual SES and leads to differences in SES and achievement between schools. On the other hand, an increase in individual SES increases the academic achievement of female students more, while male students attending schools with higher SES levels improve their academic achievement more. Increasing the individual or school SES of migrant students helps them reduce their academic achievement gaps. Moreover, the relationship between school SES and achievement was stronger in densely populated cities and large metropolitan areas. In line with the research findings, several recommendations are made to researchers and policy makers.

Deletion of the scavenger receptor Scarb1 in osteoblast progenitors and myeloid cells does not affect bone mass

The scavenger receptor class B member 1 (SCARB1), encoded by Scarb1 , is a cell surface receptor for high density lipoproteins, low density lipoproteins (LDL), oxidized LDL (OxLDL), and phosphocholine-containing oxidized phospholipids (PC-OxPLs). Scarb1 is expressed in multiple cell types, including osteoblasts and macrophages. PC-OxPLs, present on OxLDL and apoptotic cells, adversely affect bone metabolism. Overexpression of E06 IgM – a natural antibody that recognizes PC-OxPLs– increases cancellous and cortical bone at 6 months of age in both sexes and protects against age- and high fat diet- induced bone loss, by increasing bone formation. We have reported that SCARB1 is the most abundant scavenger receptor for OxPLs in osteoblastic cells, and osteoblasts derived from Scarb1 knockout mice ( Scarb1 KO) are protected from the pro-apoptotic and anti-differentiating effects of OxLDL. Skeletal analysis of Scarb1 KO mice produced contradictory results, with some studies reporting elevated bone mass and others reporting low bone mass. To clarify if Scarb1 mediates the negative effects of PC-OxPLs in bone, we deleted it in osteoblast lineage cells using Osx1-Cre transgenic mice. Bone mineral density (BMD) measurements and micro-CT analysis of cancellous and cortical bone at 6 months of age did not reveal any differences between Scarb1 ΔOSX-l mice and their wild-type (WT), Osx1-Cre, or Scarb1 fl/fl littermate controls. We then investigated whether PC-OxPLs could exert their anti-osteogenic effects via activation of SCARB1 in myeloid cells by deleting Scarb1 in LysM-Cre expressing cells. BMD measurements and micro-CT analysis at 6 months of age did not show any differences between Scarb1 ΔLysM mice and their WT, LysM-Cre, or Scarb1 fl/fl controls. Based on this evidence, we conclude that the adverse skeletal effects of PC-OxPLs in adult mice are not mediated by Scarb1 expressed in osteoblast lineage cells or myeloid cells.

In vitro toxicity assessment of uranium particulates on different human lung epithelial cell models

Inhalation of uranium aerosols produced via human activities such as mining can pose a threat to human respiratory systems. Uranium oxide particulates emit short-range alpha particles that elicit DNA and direct damage, beyond associated physiochemical heavy-metal toxicity, to internal epithelial tissues. The availability of reliable in vitro models to study radiation exposure can greatly enhance our ability to understand and combat the biological impacts of exposure. However, the toxicological effects of alpha emissions and/or the oxidation states of uranium particulates vary across different human lung epithelial cell models and have not been systematically compared. We have endeavored to address this limitation by comparing impacts in three different human lung cell models: primary human bronchial and tracheal epithelial cells, primary human small airway epithelial cells, and human adenocarcinoma alveolar basal epithelial cells. Other studies have mainly investigated the toxicity of depleted uranium. Here, we compared the exposure of uranium oxide particulates (U 3 O 8 and UO 3 ) of different enrichment states on the chosen cell systems. Each cell model was exposed to 0.1, 1, 10, 50, 100, and 500 µg/mL of depleted U 3 O 8 , highly-enriched U 3 O 8 , and natural UO 3 particulates for 24 hours in submerged monolayer cultures. We compared viability and superoxide dismutase activity results across cell lines and uranium enrichment/ oxidative states. The results showed that 1) the oxide state of the particulates affected cell viability, implying that uranium’s different oxidation states contribute to different toxicological responses, and 2) each cell model reacts differently when exposed to uranium oxides, which may provide insights into the mechanistic processes associated with the exposure of radiological particulates on different biological systems. For instance, increased uranium enrichment corresponds to increased toxicity for the primary cells, but not for the immortalized cells. Our study shows that a holistic approach that incorporates similarities between model systems and types of radionuclides is required to truly develop empirical solutions for radiation exposure.

Damascus-style hierarchical microstructures enable a strong and ductile medium-entropy alloy

Biochemical and molecular assessment of the superoxide dismutase (SOD) antioxidant enzyme regulation within Medicago truncatula root in response to iron deficiency

Iron (Fe) deficiency is a major nutritional stress affecting plant growth and metabolism. This study was conducted on three Medicago truncatula genotypes (TN8.20 and A17 as tolerant and T1.11 as a sensible genotype) cultivated in optimal and Fe-deficient conditions. Assessment of Fe deficiency effects was performed on some physiological and biochemical parameters with a particular focus on superoxide dismutase (SOD) activities and genes expression in roots. Our data showed that the sensitive genotype TN1.11 was more affected by Fe starvation compared to A17 and TN8.20. Overall, the relatively higher tolerance of A17 and TN8.20 to Fe deficiency was positively correlated to their ability to maintain higher plant biomass, Fe content, Fe use efficiency, Cu and Zn contents in roots. The oxidative stress associated with Fe-deficiency was evidenced by increased roots hydrogen peroxide (H 2 O 2 ) levels, especially in TN1.11 genotype. In contrast, Assessment of SOD activity in roots revealed a significant increase in Cu/ZnSOD and MnSOD activities under Fe-deficient conditions, particularly in TN8.20. Gene expression analysis showed differential regulation of FeSOD , Cu/ZnSOD and CHSOD genes in response to Fe deficiency. Notably, TN8.20 exhibited upregulation of Cu/ZnSOD and down regulation of CHSOD under Fe-deficient conditions. TN8.20, which showed the highest SOD activities and gene expression levels, was identified as the most tolerant genotype. These findings highlight the physiological and molecular responses of Medicago truncatula to Fe deficiency and emphasize the comparative leaf-root analyses, revealing that SOD related genes in roots may serve as useful molecular markers for selecting Fe-deficiency-tolerant genotypes to cope with oxidative stress and nutrient imbalances.