Browse Articles

DG-LLM: Decomposition-based dynamic graph adaptation of large language models for spatiotemporal traffic forecasting

Traffic forecasting plays a critical role in the field of urban planning. Yet, existing methods struggle with modeling complicated spatiotemporal dependencies and capturing long-term patterns due to their multiscale nature. In this paper, we present a novel framework named DG-LLM that leverages the advantages of decomposed temporal representations and adaptive spatial connectivity to model spatiotemporal dependencies. In this framework, traffic signals are decomposed into intrinsic modes, and dynamic graphs are learned for each mode to represent the spatial dependencies. These representations are then incorporated with pre-trained Large Language Models for effective long-range temporal dependency modeling. We conducted comprehensive experiments across six real-world traffic datasets spanning urban mobility systems and highway traffic networks and evaluated short- and long-term forecasting. Experimental results demonstrate that our framework provides significant improvements over state-of-the-art approaches, including benchmark graph- and LLM-based spatiotemporal forecasting models, even in long-term forecasting scenarios with severe temporal instability. Our model outperforms other methods by achieving 13 − 19 % improvements in MAE and 19 − 25 % in RMSE across all six benchmarks compared with baseline approaches. Additional analyses, including ablation studies, robustness to missing data, and zero-shot cross-dataset evaluation, further validate the effectiveness and generalization capability of the proposed framework.

KG-EIGNN: a knowledge graph based graph neural network for fault detection in industrial systems

Distinct endometrial protein profiles in spontaneous and stimulated cycles in women with poor ovarian response: A prospective case-crossover clinical trial

The window of implantation is a critical period for embryo implantation. Ovarian stimulation can disrupt endometrial receptivity, potentially through altered gene expression and downstream protein profiles. However, the impact on the endometrial proteome remains underexplored. Identifying biomarkers of endometrial receptivity may provide an opportunity to develop targeted interventions aimed at improving implantation outcomes. This prospective, case-crossover, open-label study was conducted at the Department of Human Reproduction, Division of Obstetrics and Gynecology, University Medical Centre Ljubljana, Slovenia, from September 2023 to June 2024. The study included 15 women aged <43 years with primary infertility and poor ovarian response. Endometrial samples were collected using a pipelle biopsy during the window of implantation in spontaneous and stimulated cycles and analyzed using protein microarrays targeting 1,466 proteins. Differential protein abundance was assessed using a multi-factorial linear model, including patient-specific effects as an additional factor to account for the paired case-crossover design. Effect sizes are reported as log2-fold changes with corresponding 95% confidence intervals. Differential abundance was defined a priori as |log2FC| > 0.5 with FDR-adjusted p-value < 0.05. Comparison of endometrial samples from spontaneous and stimulated cycles revealed 114 antibodies with differential abundance. Key proteins were associated with immune response (IL-8, proteins S100-A8 and S100-A9, CAMP) and extracellular matrix remodeling (MMP-9). Exploratory KEGG pathway mapping suggested involvement of immune and inflammatory pathways, including cytokine–cytokine receptor interaction and IL-17 signaling. Cluster analysis demonstrated distinct proteomic patterns, with all stimulated-cycle samples showing alterations and a subset of stimulated-cycle samples (40%) exhibiting more pronounced changes. The findings indicate that ovarian stimulation is associated with measurable alterations in the endometrial proteomic profile during the window of implantation. These changes may be relevant to biological pathways involved in endometrial receptivity and implantation. Further studies in larger cohorts are needed to validate the identified candidate markers and determine their clinical relevance for implantation outcomes. Trial registration: ClinicalTrials.gov NCT06804174.

Possible high-temperature superconductivity in Li$$_{2}$$CuH$$_{6}$$ at ambient pressure

Experiences shaping research career intention among Black, Hispanic, and Indigenous-identifying first-year allopathic medical students in the United States: A qualitative study

Objective To examine the early experiences influencing research career intentions (RCI) among MD students from racial and ethnic backgrounds underrepresented in medicine (URiM). Methods We conducted semi-structured, in-depth interviews with 31 first-year URiM medical students from MD-granting programs across the US to examine student-reported experiences influencing RCI. Results Participants were first-year medical students (N = 31; mean age 24.8 ± 2.6 years; 64.5% female) identifying as Black (38.7%), Hispanic (32.3%), or Multiracial (29%). Four themes were identified: (1) structured premedical research exposure was described as pivotal to developing early research engagement and interest in research careers; (2) research orientations reflected a commitment to using research as a vehicle for social justice and community impact; (3) high-quality research mentorship was characterized by authentic relational investment, skill development, and the distinct value of racial and ethnic identity-concordant role models; and (4) the research arms race for residency placement was described as amplifying systemic inequities that constrained students’ research engagement. Across themes, students described tensions between academic research culture and their personal values, including a desire to advance equity and contribute meaningfully to science. For some, this misalignment made research feel less purposeful or personally aligned. Conclusions Medical training programs seeking to support URiM students’ RCI should invest in structured premedical research programs and expand access to research mentorship that is both relationally invested and identity concordant. Efforts to cultivate sustained engagement should address publication pressures tied to residency competitiveness, which amplify structural barriers and misalign with students’ motivations for pursuing research. Broadening definitions of scholarly contribution and fostering research environments that affirm students’ values may be critical to building a robust physician-scientist workforce.

Monkey jumping optimization: a tree-branch-inspired metaheuristic for global search

Indoor methane consistently above outdoor levels in homes with natural gas service

Background Methane leaks across the natural gas process chain, including in homes. To date, no studies have described how common it is to have elevated methane in homes served by gas, in comparison to homes without gas. Methods In this study of homes in Massachusetts and Rhode Island, we utilized Cavity Ringdown Spectrometry to measure methane concentrations in outdoor air, and at mid-floor-level indoor air in basements and first, second, and third floors. We recruited a total of 195 homes in urban and rural areas, 175 of which had gas service and 20 of which did not. Results Indoor [CH 4 ] in households with gas service was elevated over outdoor [CH 4 ], averaging 1.45 parts per million (ppm) elevation over outdoor ambient [CH 4 ] (p < 0.0001–0.0068), and up to 38.2 ppm above outdoor ambient [CH 4 ]. Ninety-three percent of homes with gas showed higher median indoor [CH 4 ] than the median [CH 4 ] in non-gas homes. By contrast, indoor [CH 4 ] in gas-free homes did not differ from outdoor conditions (p > 0.10), except marginally on the first floor (0.10 ppm elevation; p = 0.04). In 91% of a subset of homes investigated, leaks from gas equipment were confirmed. Conclusions Elevated [CH 4 ] is common in homes served with gas. Gas leaks and incomplete combustion were identified as sources of elevated [CH 4 ]. There was no relationship between indoor [CH 4 ] and home age or square footage; residents shouldn’t assume that newer homes are less prone to indoor gas leaks. The majority of gas in the United States comes from hydraulically fractured gas containing carcinogenic co-pollutants. It is not well understood how consistent low-dose exposure to gas co-pollutants like mercaptans and benzene affects health. Additional studies could clarify any differences in health outcomes for people living in homes serviced by gas and those who don’t use gas.

Adaptive SMC with direct dual inertial matrix compensation for spacecraft attitude-orbit dynamics

Contracting the gap: The effects of identity gaps on the psychological adaptation of Taiwanese university students

With the rapid internationalization of higher education, patterns of student mobility have moved beyond the traditional intercultural model of East-to-West migration. Mainland China has become a major destination for international students, and the number of Taiwanese university students studying there has steadily increased recently. This trend underscores a significant research gap regarding the psychological adaptation of students engaged in intra-cultural mobility within Eastern cultural contexts. This study extends identity gap research, a core concept of the Communication Theory of Identity (CTI), to a less explored intra-cultural educational mobility framework. It empirically examines the associations among Personal Enacted Identity Gap (PEIG), Personal Relational Identity Gap (PRIG), and Personal Communal Identity Gap (PCIG), as well as their associations with the psychological adaptation of Taiwanese university students in mainland China, with the aim of identifying the identity construct most consistently associated with psychological adaptation. A cross-sectional survey was conducted with 425 Taiwanese university students in Fujian Province. Data were analyzed using partial least squares structural equation modeling (PLS-SEM), two alternative models, and PLS-IPMA. The findings show that identity gaps are negatively associated with psychological outcomes in an Eastern cultural context, and that the PCIG is the construct most consistently negatively associated with psychological adaptation. These findings extend the understanding and applicability of the CTI and highlight the cultural sensitivity of identity gaps across different contexts. This study also provides important implications for policymakers and educational administrators seeking to design and implement strategies to promote the psychological adaptation of Taiwanese university students in mainland China.

Circular mixture model-based characterization of time-of-day seizure patterns during presurgical monitoring

Abstract Epilepsy is one of the most prevalent neurological diseases. Despite advancements in treatment, Drug-resistant epilepsy continues to pose significant challenges. Identifying temporal patterns in seizure occurrence may provide insight into the mechanisms underlying seizure timing. This study analyzes seizure patterns in patients undergoing presurgical evaluation using the long-term electroencephalogram EPILEPSIAE database. After applying exclusion criteria, 280 patients were analyzed, encompassing 1813 annotated seizure onsets. Seizure timing was treated as a circular variable, and each patient’s data were fit to a von Mises Mixture Model, with each component representing a cluster of seizure onset times. Clustering strength was evaluated using the Bayesian Information Criterion by comparing each model against a uniform one. To address the increased false positive rate associated with small sample sizes, primary results are reported for patients with more than 8 seizures. In this subgroup, model-supported evidence of time-of-day seizure onset clustering was found in 30% of patients. Results from patients with fewer seizures are reported separately as exploratory analyses due to reduced model stability at low sample sizes. These findings demonstrate that, for some patients, structured time-of-day patterns can be detected even during presurgical monitoring, where medication tapering, sleep disruption, and the hospital environment may perturb seizure timing, suggesting that rhythmic influences on seizure occurrence may persist despite substantial environmental perturbations.

Long-term changes in the Juvenile Sockeye Salmon Rearing Capacity of the Chignik Lakes Watershed

Freshwater ecosystems respond rapidly to perturbations in climate, geomorphology, and population abundances. For migratory species in interconnected habitat networks, local habitat conditions can control the productivity of individual populations. Asynchronous variation in habitat quality can simultaneously stabilize ecological processes at broad scales but also complicate understanding of ecosystem dynamics. We investigated habitat-specific trends in indicators of the rearing capacity for juvenile sockeye salmon in a remote watershed in Alaska over the last ~60 years. The motivation of this effort was to understand if the collapse of the local salmon fishery in 2018 could be traced to changes in habitat quality within the nursery watershed. Our analyses describe high variability in the habitat conditions across both spatial and temporal scales, yet do not suggest a decline in the overall sockeye salmon rearing capacity of the watershed. We observed increasing maximum water temperatures in a shallow lake but more stable conditions in a deep lake, an improvement in zooplankton prey resources in the deep lake, and increased juvenile sockeye salmon growth rates throughout the watershed. Although we detected no long-term decline in rearing habitat quality, there was a decrease in juvenile sockeye salmon abundance from 2013–2016, suggesting high early life stage mortality prior to the period of juvenile rearing leading up to the fishery collapse.

Quantitative assessment of supply chain performance drivers through structural equation modelling approach

Psychosocial interventions for adolescents in forced migration: A scoping review protocol

Background Adolescents account for 40% of forcibly displaced people worldwide and are often the most vulnerable to mental health challenges. Although psychosocial interventions are becoming popular, there remains a significant gap in the application of psychosocial interventions for adolescents in forced migration context. Mapping the existing literature is necessary to help guide the use of psychosocial interventions for this population. Objective To systematically map and synthesise the scope of existing literature related to psychosocial interventions for adolescents in forced migration situations. This review will inform the policy and practice. Inclusion criteria This review will include empirical studies using qualitative, quantitative, or mixed methods approaches. The review will include any study where adolescents aged 10–19 years in forced migration settings receive psychosocial intervention with the purpose of improving psychological well-being, mental health, or psychosocial functioning. The study population will include adolescent refugees, asylum seekers, internally displaced persons, and other adolescents who have been compelled to leave their homes due to conflict, persecution, natural disasters, or other threats to safety and security. Studies will include both high-income and low- and middle-income countries. Only studies published from 2000 to the present will be included in the review. Method The review will follow the Joanna Briggs Institute methodology and PRISMA-ScR reporting guidelines. It will include a comprehensive search of published studies in databases and grey literature. Two independent reviewers will evaluate all abstracts and full articles that are written in English. We will identify concepts from psychosocial interventions, categorise accepted studies according to characteristics, describe implementation factors, and report outcomes of selected studies. Scoping review registration This study protocol is registered with Open Science Framework (OSF) with DOI 10.17605/OSF.IO/ADF92 .

Components wear detection by piezo-acoustic sensors and influence on the material in progressive stamping

Abstract Progressive stamping is widely used for high-volume sheet metal production, where tool condition directly affects stability and product quality. Tool wear, such as punch and die edge rounding, impacts the piercing mechanism and increases process forces and burr formation, and causes part edge deterioration. Therefore, accurate real-time wear monitoring is essential to prevent failures and reduce downtime. This study explores piezo-acoustic sensors for real-time monitoring of punch wear. Experiments on DC01 steel included static and dynamic tests with predefined tool wear. Sensor signals and force data were continuously analyzed. Results show that increased wear leads to higher peak load and delayed material fracture, indicating a shift from shear-dominated piercing to deformation-dominant material separation. Microstructural analysis confirmed that wear significantly alters fracture morphology and subsurface deformation, impacting part quality. Research shows that piezo–acoustic sensing yields the peak load value for punch wear, making it a powerful and useful measure in industry. It provides an effective solution for real-time tool-condition monitoring and predictive maintenance in progressive stamping operations.

Advantages and disadvantages of self-determination support strategies for people with severe or profound intellectual and multiple disabilities: A Delphi study

Background The specific advantages and disadvantages of strategies that caregivers use to support self-determination in individuals with profound intellectual and multiple disabilities need to be known. This study explored consensus on advantages and disadvantages of support strategies from the perspective of users. Method A four-round Delphi study was conducted with 10 relatives and 13 professionals experienced in self-determination support. In Round 1, qualitative data were collected and thematically analysed. Subsequent rounds used Likert scales and participant feedback to explore consensus. Results Thematic analysis of Round 1 produced 79 themes, on which 72 consensus was reached across later rounds. On 63 themes, consensus was found across both groups. Users reported advantages such as mutual trust, reciprocal learning, and a more unified and efficient approaches to support self-determination. For individuals with disabilities, advantages included improved health, participation, and increased influence over their lives. However, disadvantages included preconditions that are often unmet in practice, such as consistent use of communication aids, sufficient time, and stable staffing. Conclusion Promoting self-determination in this population requires attention to their interactions with support networks and the quality of collaboration. Each support strategy has specific advantages and disadvantages that should be examined in light of the diversity within this population and their networks. Future research and practice could address identified barriers and build on advantages of support strategies by evaluating their applicability and relevance in individual cases.

Optimizing green-adaptive green infrastructure in response to extreme climate changes: insights from a hybrid SEM-ANN method

Research on railway emergency resource scheduling strategy under multiple uncertainty coupling

To address the multiple sources of uncertainty in railway emergency resource scheduling, this study uses interval numbers to characterize uncertainty in scheduling time, models resource demand as a fuzzy-random variable, and integrates fuzzy credibility constraints with interval programming within a unified optimization framework. Based on interval fuzzy credibility-constrained programming, an emergency resource scheduling model is developed to minimize both scheduling time and scheduling cost. Given the characteristics of the model constructed in this paper, an improved solution strategy based on the VEPSO algorithm is designed. By using the mean of existing feasible solutions to repair infeasible solutions generated during the optimization process, the proposed method ensures that all particles satisfy the constraints. This approach effectively addresses the complex and discontinuous Pareto front characteristics inherent in multi-objective programming involving multiple uncertain variables, enabling rapid search and selection of Pareto-optimal solutions. The results of the case study indicate that the VEPSO algorithm outperforms PSO, NSGA-II, VEGA, and MOEA/D in terms of convergence efficiency and solution quality. Its average number of convergence iterations is only 385, which is approximately 26.0% lower than that of MOEA/D, with an average runtime of 273 seconds, a convergence success rate of 93.3%, and a total demand fulfillment credibility of 0.983. By introducing a satisfaction evaluation method, this study quantifies the degree of alignment between resource allocation and actual demand under different confidence levels across various accident scenarios, revealing the distribution patterns by which both factors jointly determine resource satisfaction. This provides a quantitative basis for dynamically adjusting emergency resource scheduling strategies. Furthermore, a sensitivity analysis of resource demand was conducted at a fixed confidence level. The results indicate that when the demand scaling factor k increases from 1.0 to 1.2, the rate of change in total scheduling cost is 16.0529%, and the rate of change in total scheduling time is 50%. The model exhibits a stable and approximately linear response, indicating good robustness. This study provides quantifiable decision support for railway emergency resource scheduling in environments with multiple uncertainties, offering both theoretical value and practical significance.

Synergistic activation of dendritic cells by Lycium barbarum polysaccharide and tumor endothelial progenitor cell lysate enhances anti-tumor responses

Shared drone route scheduling optimization

Shared passenger-carrying unmanned aerial vehicle (UAV) systems offer a promising solution for urban air mobility, yet their real-time scheduling in multi-airport environments remains challenging due to heterogeneous passenger demands, model compatibility constraints, and the need to balance efficiency with service quality. This study addresses these challenges by proposing a shared UAV route scheduling optimization model for multi-airport, multi-station systems. The model integrates passenger order characteristics (origin, destination, party size, preferred UAV type, and desired arrival time) with UAV operational parameters (current location, remaining endurance, seating capacity, speed, and model), and incorporates a quadratic soft time window mechanism to flexibly manage arrival deviations while minimizing total system navigation time. To solve this discrete combinatorial optimization problem, we develop a self-learning Ant-Lion Optimizer (SLALO) with natural number encoding that dynamically tracks seat availability and updates route assignments in real-time. Simulation results demonstrate that the proposed approach achieves a 27.3% reduction in total navigation time compared to non-pooling baselines, with an average UAV utilization rate of 78.6%. Comparative analysis against Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and standard Ant-Lion Optimizer (ALO) shows that SLALO achieves superior convergence speed and solution quality. The proposed framework is particularly relevant for border regions with limited ground infrastructure, where intelligent shared mobility systems can serve as a critical driver for cultivating new quality productive forces and supporting regional economic integration. These findings highlight the potential of shared UAV systems to improve urban air mobility efficiency, reduce operational costs, and enhance passenger satisfaction.

Metaheuristic-optimized bell-shaped slotted pentaband patch antenna for C, X, Ku, and K band applications

Abstract The current study proposes a compact metaheuristic optimization-driven design of the slotted pentaband patch antenna with a bell-shaped patch, operating in the C, X, Ku, and K bands for satellite telemetry, radar sensing, and wideband transceivers. The antenna structure comprises a bell-shaped patch with C-slots, tilted vertical edges, and three semicircular perturbations on the horizontal edge of the patch to excite several resonant modes, mainly TM10, TM11, and TM20, with uniform surface current distribution. The antenna has five resonant frequencies at 5.2, 8.4, 10.4, 12, and 19.2 GHz with a reflection coefficient of less than -15 dB and maximum antenna gains of 8.83 dB, 4.96 dB, 7.87 dB, 6.13 dB, and 6.37 dB at the corresponding resonant frequencies. The antenna also has maximum radiation and total efficiencies of at least 85% in the C and X bands and above 95% in the Ku and K bands. In addition, the antenna performance is improved using five different metaheuristic optimization techniques, and the optimal results are obtained using the multi-objective optimization technique. Among the optimization techniques, the performance of the Differential Evolution (DE) algorithm is found to be the best, with improvements of up to 4.0% in S11, 2.2% reduction in VSWR, and improvements in antenna gains and efficiencies of up to 0.7%. The proposed antenna design using the electromagnetic simulator, circuit simulator, and optimization tools provides the best platform for the proposed antenna and is further validated experimentally.