Search Results (2,002)

Background: The optimal sequencing of regorafenib and trifluridine/tipiracil (FTD/TPI) in refractory metastatic colorectal cancer (mCRC) remains uncertain, particularly in Asian populations. Methods: We retrospectively analyzed 110 patients with mCRC who sequentially received both agents between 2011 and 2025. Patients were categorized into regorafenib followed by FTD/TPI (Rego → FTD/TPI, n = 88) and FTD/TPI followed by regorafenib (FTD/TPI → Rego, n = 22). Co-primary endpoints were time to treatment discontinuation (TTD) and overall survival (OS). Propensity score-based weighting methods, including stabilized inverse probability of treatment weighting (primary analysis), were used to adjust for baseline imbalances. Multivariable Cox regression was performed as a sensitivity analysis. Results: No statistically significant differences were observed between treatment sequences. In the primary analysis, the hazard ratio (HR) for TTD was 1.01 (95% CI 0.71–1.43), and for OS was 1.19 (95% CI 0.67–2.12), with FTD/TPI → Rego as reference. Median TTD was 6.8 versus 8.9 months, and median OS was 14.6 versus 20.2 months, respectively. Conclusions: Clinical outcomes were comparable regardless of treatment order, supporting individualized sequencing decisions in refractory mCRC. Full article

The increasing complexity and variability of operating conditions, along with the occurrence of low-probability cascading failures, imposes more stringent requirements on data-driven intelligent methods for power system stability analysis. This paper proposes an intelligent enhancement approach for transient voltage stability assessment in modern power systems, considering improved conditional generative adversarial network (CGAN)-based sample balancing. Firstly, an improved CGAN incorporating an enhanced feature-distance metric is developed to accurately capture the distribution characteristics of real samples, effectively alleviating training issues such as gradient vanishing and mode collapse during adversarial learning. Secondly, an intelligent sample enhancement method for transient voltage stability is established based on the improved CGAN, which effectively complements the initial dataset and ensures the predictive performance of intelligent models under extreme operating conditions. Finally, a transient voltage stability assessment framework integrating a convolutional neural network and a transformer is proposed to enable efficient extraction of low-dimensional features and achieve accurate evaluation of transient voltage stability states. Full article

Heavy metals in lake sediments represent typical persistent contaminants characterized by recalcitrance, bioaccumulation potential, and delayed toxic effects, thereby exerting sustained adverse impacts on lacustrine ecosystem stability and human health. Liuye Lake is a representative small-to-medium urban lake impacted by ambient domestic sewage discharge and agricultural non-point source pollution, with documented nitrogen and phosphorus enrichment. However, the contamination profile of heavy metals in its surface sediments has not been systematically investigated to date. In this work, surface sediment samples were collected from Liuye Lake, and nine heavy metal elements (As, Cd, Cr, Cu, Hg, Mn, Ni, Pb, Zn) were determined. An integrated approach incorporating Monte Carlo simulation, the geo-accumulation index (I_geo), and the enrichment factor (EF) method was employed to assess the ecological risk and human health risk imposed by these metals. The results revealed the following: (1) Average concentrations of eight heavy metals exceeded the background values of the Dongting Lake water system, with the exception of As, and Hg displayed potential localized anomalies. (2) Surface sediments were collectively categorized as slightly contaminated, with Hg identified as the primary pollutant, followed by minor contamination of Mn, Cr, and Ni; Monte Carlo simulation further suggested a probable risk that Mn contamination could progress to moderate levels. (3) All heavy metals posed low potential ecological risk, with an overall potential ecological risk index (RI) of 62.71, where Cd, Hg, and As were the dominant contributors. (4) Both non-carcinogenic and carcinogenic risks were generally within acceptable limits, whereas children exhibited higher non-carcinogenic susceptibility relative to adults; As and Mn were the leading contributors to non-carcinogenic risk, while Cr and As dominated carcinogenic risk. This study offers a scientific foundation for the prevention and control of heavy metal pollution and the ecological management of urban lakes. Full article

Background: Short-term facilitation (STF) is a key form of synaptic plasticity driven by activity-dependent increases in presynaptic release probability. However, estimating core synaptic parameters—quantal size ($q$), vesicle pool size ($N$), and release probability ($p_i$)—remains challenging due to nonlinear dynamics and unobservable presynaptic states, limiting the applicability of conventional methods. Methods: We developed a fast analytical framework based on first- and second-order statistical moments of evoked EPSCs, including mean, variance, and cross-stimulus covariance. By constructing composite moment relationships, latent variables were algebraically eliminated, yielding closed-form estimators of synaptic parameters. To improve robustness under strong facilitation, a Tsodyks–Markram (T–M) model-based calibration step was introduced to refine $N$ and $p_i$ using the estimated $q$ as a constraint. Results: Applied to hippocampal CA3–CA1 synapses, the method produced accurate and stable estimates of q across varying noise and sampling conditions. Incorporation of cross-stimulus covariance enabled effective characterization of structured variability that is neglected in classical approaches. While direct estimates of $N$ and $p_i$ showed dispersion, T–M calibration significantly improved stability and physiological consistency. Compared with mean–variance analysis, the proposed method achieved superior performance under facilitating conditions. Conclusions: This hybrid framework enables rapid and reliable estimation of synaptic parameters in STF synapses by exploiting second-order statistical structure. It provides a practical tool for investigating presynaptic mechanisms and may facilitate quantitative studies of synaptic dysfunction in neurological and psychiatric disorders. Full article

Greenhouse cultivation enables year-round vegetable production and high yields through precise environmental regulation. Yet, the same stable microclimate that promotes crop growth also favors the proliferation of pests and diseases. This review synthesizes current knowledge on how greenhouse climate variables govern pest and disease epidemiology in tomato, cucumber, and sweet pepper. Only greenhouse-based studies were included to ensure direct relevance to protected horticulture. Microclimatic stability determines infection probability, vector behavior, and host susceptibility. Warm, humid conditions promote fungal and bacterial pathogens, whereas dry, high vapor pressure deficit (VPD) environments favor mites and thrips and enhance virus transmission. Species-specific traits further modulate vulnerability. Tomato is dominated by virus–bacterium complexes and foliar/stem fungal diseases, cucumber by phytopathogenic fungi favored by high relative humidity (RH) and soilborne pathogens, and sweet pepper by virus–vector systems and long-cycle fungal infections. Temperature exerts the strongest influence, while RH and VPD jointly regulate surface moisture and vector activity. Light intensity and spectral composition also affect pest orientation and fungal sporulation. Integrating environmental sensing, biological control, and adaptive climate regulation offers a pathway toward preventive, climate-smart Integrated Pest Management (IPM). The review highlights the emerging role of climate-informed decision-support systems (DSSs) and the need for greenhouse-specific datasets to improve pest and disease forecasting. Full article

Objectives: To evaluate extrusion-free survival following glaucoma drainage device (GDD) surgery using EverPatch Plus^® (EPP) and to compare outcomes with conventional scleral patch grafts using propensity score-based survival analysis. Methods: This retrospective case series included 19 eyes that underwent GDD implantation with EPP and 105 control eyes that received conventional scleral patch grafts. To adjust for baseline differences between groups, a propensity score for EPP use was estimated using multivariable logistic regression incorporating age, neovascular glaucoma, prior glaucoma surgery, preoperative intraocular pressure, number of glaucoma medications, quadrant of patch placement, and insertion site. Stabilized inverse probability of treatment weighting was applied. Because follow-up in the EPP group did not exceed 12 months, all survival analyses were performed with administrative censoring at 12 months. Extrusion-free survival was evaluated using Kaplan–Meier analysis and Cox proportional hazards modeling. Results: Within 12 months, patch extrusion occurred in 3 of 19 eyes in the EPP group and in 12 of 105 eyes in the scleral patch graft group. After inverse probability weighting, estimated 12-month extrusion-free survival was 83.5% in the EPP group and 88.4% in the scleral patch graft group, indicating no statistically significant difference between groups (log-rank test, p = 0.498). In an inverse probability-weighted Cox model, EPP use was not significantly associated with extrusion risk (hazard ratio ≈ 1.3; 95% confidence interval ≈ 0.4–4.0). Conclusions: After adjustment for baseline covariates and restriction of follow-up to 12 months, extrusion-free survival following glaucoma drainage device surgery using EPP was comparable to that achieved with conventional scleral patch grafts. Full article

Against the backdrop of climate change, forest fires increasingly undermine ecosystem stability and reshape species distributions in Heilongjiang Province. Therefore, quantifying the drivers of fire occurrence and conducting long-term fire risk forecasting holds critical value for regional ecological security. Centered on the forested regions of Heilongjiang Province, this study systematically assessed the relative contributions of multi-source factors—including topography, vegetation, and meteorological conditions—to fire occurrence and compared the predictive performance of three models: Deep Neural Network with Residual Connections (ResDNN), Artificial Neural Network (ANN), and Support Vector Machine (SVM). Modeling results based on historical fire records indicated that the ResDNN model achieved the highest accuracy (85.6%). Owing to its robust nonlinear mapping capability, it performed better in capturing complex feature interactions than ANN and SVM. These results demonstrate its strong applicability to forest fire prediction in Heilongjiang Province. Building on these findings, the study employed the best-performing ResDNN model in conjunction with CMIP6 multi-model climate projections to simulate and map the spatiotemporal probability of forest fire occurrence from 2030 to 2070. The results provide an intuitive representation of long-term fire-risk trajectories under future climate scenarios and offer scientific support for regional fire prevention, monitoring, early-warning systems, and forest management under climate change. Full article

This research draws from participant interviews at baseline, midpoint, and conclusion of a two-year Basic Income program designed to lift 76 families, with at least one child under the age of six, above the California poverty line by supplying a guaranteed monthly cash stipend (average: $1289 per month). Theoretically, we bring the Family Stress Model into the conversation with the Theory of Change that underpins Guaranteed Income programming, namely that freedom and choice empower families to mitigate stress, manage funding, and better navigate the multifactorial aspects of living in poverty. Participants report that the Basic Income program offered a much-appreciated reprieve from poverty and reported using the funds to stabilize their housing and support the health and development of themselves and their children. Participants also highlighted how guaranteed cash programming can pair with traditional social service case management to better benefit recipients. Full article

The accurate inference of hidden states from non-stationary physiological signals remains a significant challenge in stochastic process modeling. This paper proposes an Adaptive Sticky Hidden Markov Model (Sticky-HMM) framework designed to enhance the robustness of state decoding in noisy environments. To address the “state-flickering” issue inherent in traditional HMMs, we incorporate a “Sticky” parameter into the transition matrix, imposing a temporal penalty on spurious state switching to maintain continuity. Furthermore, we introduce a Dynamic Prior Strategy that adaptively calibrates self-transition probabilities by mapping frequency-domain features of the observed sequence to the model’s parameter space. The proposed decoding process employs a two-pass refinement strategy and the Viterbi algorithm in the logarithmic domain to ensure numerical stability. The model’s efficacy was validated using a high-fidelity dataset of simulated apnea events. This work provides a computationally efficient and mathematically rigorous approach that demonstrates strong potential for long-term respiratory health monitoring. Full article

In this paper, we investigate the spectrum distribution and asymptotic behavior of an M/$G^B$/1 bulk service queueing system. In this system, the server processes customers in batches of a fixed maximum capacity B, and the time required to serve a batch is governed by a general distribution with a service rate function $\eta (·)$, which determines the instantaneous probability of service completion. The system dynamics are described by an infinite set of partial integro-differential equations. First, by introducing the probability generating function and employing Greiner’s boundary perturbation method, we establish that the time-dependent solution (TDS) of the system converges strongly to its steady-state solution (SSS) in the natural Banach state space. To this end, when the service rate $\eta (·)$ is a bounded function, we prove that zero is an eigenvalue of both the system operator and its adjoint operator, with geometric multiplicity one. Moreover, we show that every point on the imaginary axis except zero belongs to the resolvent set of the system operator. Second, we analyze the spectrum of the system operator on the left real axis. When the service rate $\eta (·)$ is constant and the fixed maximum capacity B equals 2, we apply Jury’s stability criterion for cubic equations to demonstrate that the system operator possesses an uncountably infinite number of eigenvalues located on the negative real axis. Additionally, we prove that an open interval near zero on the left real axis is not part of the point spectrum of the system operator. Consequently, these results imply that the semigroup generated by the system operator is not compact, eventually compact, quasi-compact, or essentially compact. Full article

The arithmetic optimization algorithm (AOA) is a recently proposed swarm intelligence optimizer with a simple structure and few control parameters. However, the original AOA relies on a single update mechanism, which often leads to premature convergence and limited adaptability in complex optimization problems. To address these limitations, this paper proposes a multi-strategy improved arithmetic optimization algorithm (IAOA). The proposed algorithm constructs a heterogeneous strategy pool composed of six search strategies, including arithmetic update, differential evolution operators, competitive elite learning, interpolation-based acceleration, and curriculum education learning. Furthermore, an adaptive strategy regulation mechanism based on fitness improvement contribution is introduced to dynamically adjust the selection probability of each strategy. Extensive experiments conducted on the CEC2017 and CEC2022 benchmark suites demonstrate that IAOA achieves a superior optimization accuracy, convergence speed, and stability compared with several classical algorithms, recent metaheuristics, and AOA variants. Statistical tests including the Wilcoxon rank-sum test and Friedman mean rank test confirm the significance of the performance improvements. In addition, the algorithm is successfully applied to a three-dimensional path planning problem for amphibious unmanned aerial vehicles, demonstrating its effectiveness in solving complex engineering optimization problems. Full article

Rainfall-induced shallow loess landslides pose a significant threat to human life and property. Early warning and risk assessment of these landslides are critical prerequisites for engineering control and disaster loss reduction. The Transient Rainfall Infiltration and Grid-Based Regional Slope-Stability Model (TRIGRS)-Three-dimensional Slope Stability Analysis Tool (Scoops 3D) joint model can overcome the shortcomings of using a single TRIGRS model for hydrological analysis and a single Scoops 3D model for slope stability analysis. Landslide risk assessment based on expected economic loss, on the other hand, can overcome the issue of maintaining the risk level edge and sorting at the same level. In this paper, the TRIGRS model’s head pressures were put into the Scoops 3D model, with the southeast of Fangta, a town in Shaanxi province, China, as the study area. The relationship between the slope gradient and the number of grids in each stable grade was certified. The rainfall thresholds for landslides, based on both rainfall intensity and rainfall duration, were obtained by rerunning the TRIGRS-Scoops 3D joint model. The landslide range and land uses of each dangerous slope were determined by maximum likelihood classification, and then the expected economic loss was calculated. To verify the reliability of the TRIGRS-Scoops 3D joint model, the identified dangerous slopes were compared with the results from landslide susceptibility mapping. The results show that the unstable grids are concentrated within a slope gradient of 30° to 35°, and the landslide early warning levels are divided into Tier 3, Tier 2, and Tier 1 Warnings. The occurrence of shallow loess landslides is affected by both rainfall intensity and rainfall duration, and the combined effect should be considered in early warning. The distribution of both extreme susceptible grids and high susceptible grids across all 23 dangerous slopes demonstrates the reasonableness of the TRIGRS-Scoops 3D joint model. The landslide susceptible probability within some dangerous slopes exhibits spatial variability. The mapping relationship between the slope gradient and loess landslides is extremely complex. This paper can provide a theoretical basis for the early warning and risk management for rainfall-induced shallow loess landslides; the proposed method is also applicable to other regions with similar geological and meteorological conditions. Full article

Current peer-to-peer (P2P) energy trading systems face important challenges in decentralised trading environments, particularly in managing participant trustworthiness, preventing dishonest behaviour, and mitigating transaction defaults. These limitations reduce transaction reliability and weaken trust among participants in community-scale energy trading markets. Although P2P energy trading enables communities to exchange locally generated renewable energy in smart environments, existing platforms often lack effective mechanisms to regulate participant behaviour and support reliable transactions. This paper proposes RepuTrade, a blockchain-based P2P energy trading platform tailored for community-scale microgrids. The proposed framework integrates a reputation-based consensus mechanism and a dynamic collateral management scheme that is directly linked to participant reputations such that trading reliability can be strengthened through behavioural incentives. In addition, a reputation-driven matching algorithm preferentially pairs highly reputable participants to improve market stability and trust. Simulation-based evaluation, involving 200 users across 8 trading rounds, shows that the RepuTrade framework consistently achieves higher trade success rates (92–99% compared to 83–95% in the baseline) and reduces defaults by more than 40% (27–44 vs. 55–72 per run). The results further reveal a strong negative correlation between user reputation and default probability, indicating that higher reputation is associated with a lower likelihood of dishonest behaviour. Overall, under the simulated settings considered in this study, the proposed framework improves transaction reliability and execution efficiency by reducing failed trades and lowering consensus validation latency. These findings contribute to the design of trust-aware decentralised energy trading mechanisms and provide simulation-based insights for developing more reliable and transparent community-scale renewable energy markets. Full article

The stability of solar-dependent energy systems is vital for urban sustainability, but it is increasingly threatened by compound energy risks (CERs), events where low photovoltaic generation coincides with high electricity demand. This study addresses a critical knowledge gap by disentangling the co-evolving structural and magnitude drivers of these events to identify their propagation pathways and the most vulnerable districts. To achieve this, a novel hybrid framework was developed to provide a high-resolution, spatiotemporal assessment of both risk dimensions across Singapore’s 41 districts. Structural risk was mapped by integrating an undirected co-occurrence network, quantified using Mutual Information (MI), with a directed influence network derived from Bayesian Network Theory (BNT). Concurrently, magnitude risk was assessed through a copula-based analysis of joint probabilities for historical and future climate conditions, using Singapore’s new V3 dataset under multiple Shared Socioeconomic Pathways (SSPs). The results reveal a significant shift in the compound energy risk landscape. Structurally, the network of risk propagation evolves from a historically diffuse configuration to a consolidated system dominated by clusters of 8 to 9 highly interconnected districts under the SSP245 scenario. Under the high-diffusion SSP585 scenario, this evolution is expanded by the addition of 4 more districts. At the same time, the magnitude of risk intensifies across identified hotspot districts. This synthesis uncovers a critical feedback dynamic: districts such as 29, 36, and 40 not only serve as key structural hubs but also experience sharp increases in event probability, with their return periods for extreme compound events collapsing from over 50 years historically to the 10–20-year range. This forms a self-reinforcing loop of systemic vulnerability. These findings indicate that Singapore’s energy security will become increasingly exposed to climate-driven risks that propagate through this consolidated network, requiring targeted spatial adaptation to ensure long-term grid sustainability. Full article

In the context of the rapid advancement of financial technology, the issue of credit card default has become increasingly salient, emerging as one of the crucial risks that financial institutions are eagerly addressing. Traditional credit card default risk prediction models predominantly rely on statistical correlations for feature selection. This approach not only makes it challenging to uncover the genuine causal relationships between variables but also leads to limitations in prediction accuracy and interpretability. To overcome these limitations, this paper presents a novel credit card default risk prediction model that integrates causal feature screening, interaction feature construction, and interpretability enhancement. Initially, by leveraging the information value (IV) and eXtreme gradient boosting (XGBoost), we perform initial feature dimensionality reduction. Subsequently, we introduce the Peter Clark algorithm (PC) augmented with perturbation enhancement and bootstrap sampling to identify a stable set of causal features. Building on this foundation, we proceed to construct higher-order interaction features to bolster the model’s nonlinear modeling capacity. These causal features and their interaction counterparts are then fed into a variety of mainstream machine learning models for training and evaluation purposes. Furthermore, on the basis of the causal feature set identified via the PC algorithm, we construct a causal path diagram. We also incorporate the causal forest double machine learning (causal forest DML) method to estimate the causal effects of features. Additionally, we design a counterfactual explanation mechanism to aid in analyzing the direction and magnitude of the impact of variable interventions on default probability. Empirical tests conducted using four typical credit datasets reveal the following findings: (1) the introduction of causal features generally enhances the model’s performance in terms of the F1 score, area under the curve (AUC), and geometric mean (G-mean). This improvement is especially pronounced in models that are highly reliant on feature quality, such as logistic regression (LR). (2) Causal features offer significant advantages in terms of model interpretability, stability, and compliance, thereby presenting a new research paradigm for credit risk prevention and control in high-risk financial scenarios. Full article