Search Results (21,668)

Large Language Models (LLMs) are increasingly deployed as automated annotators in semantic multimedia systems, yet their reliability varies significantly across architectures. This study extends prior cross-model evaluations by benchmarking ChatGPT-5, Qwen-3, and Gemini-3-flash against human expert annotations using the HRAST hotel review dataset. We adopt a bias-by-design framework to analyze systematic divergences in sentiment, topic, and aspect labeling across real and synthetic data, while investigating the moderating effects of annotation mode. Findings reveal model-contingent polarity bias: ChatGPT-5 exhibits a pronounced neutrality bias, while Qwen-3 and Gemini-3-flash align more closely with human polarization. Agreement is substantial for concrete topics but diverges on abstract evaluative dimensions. Synthetic data consistently inflates reliability metrics while masking ambiguity. These findings highlight that annotation bias is structurally embedded in model design choices and operational conditions. Cross-architectural triangulation and mode-aware deployment strategies are recommended for robust semantic multimedia system development. Full article

Background: Hematological patients have a high risk of developing severe COVID-19 (37%). Most mRNA vaccine trials in hematological patients showed a low immunogenicity after two doses, while long-term data are scarce. Methods: In this monocentric retrospective observational study, we evaluated humoral and T cell-mediated immune responses in 230 hematological patients after three doses of the Pfizer-BioNTech mRNA COVID-19 vaccine. Patients were stratified by age, disease type/state, prior COVID-19 infection, and treatment status and regimens (anti-CD20 monoclonal antibodies, BTK and BCL-2 inhibitors, and treatment line). Antibody titer to SARS-CoV-2 was assessed by electrochemiluminescence immunoassay and T cell response by QuantiFERON interferon-γ release assay (IGRA). Data were analyzed using univariate (Fisher’s exact test) and Firth’s bias-reduced penalized-likelihood logistic regression. Results: A robust humoral response was observed with 91.55% of patients developing anti-spike antibodies (GMT 988.83 U/mL). Anti-CD20-bendamustine treatment was associated with a significantly lower antibody positivity compared to untreated subjects. Prior COVID-19 infection significantly boosted both antibody positivity (95.9% vs. 85.2%) and GMT (847.02 U/mL vs. 258.79 U/mL). Conversely, T cell response was suboptimal (36.1% positive), particularly in anti-CD20-bendamustine-treated and multi-treated patients (27.1%), but highest in those treated with BTK inhibitors (50%). Multivariable logistic regression analysis linked multiple treatments to lower T cell response. Following vaccination, 29.1% of patients contracted SARS-CoV-2, but only 0.89% developed severe COVID-19. Conclusions: Three doses of mRNA vaccine elicit a strong humoral but a low T cell response, as detected by IGRA, in hematological patients. These findings underscore the importance of completing vaccination before initiating immunosuppressive therapies. Full article

This study presents a comprehensive assessment of historical and projected variability in the wind climate over Ireland and its adjacent marine regions, including the North Atlantic Ocean, Irish Sea, and Celtic Sea. First, the long-term wind characteristics are examined using the ERA5 reanalysis dataset for the historical period (1979–2008), followed by an evaluation of five CMIP6 Global Climate Models (GCMs) to determine their performance in representing regional wind climatology. Based on spatial validation and relative bias analyses, the most suitable model is selected to investigate future wind changes under the SSP2-4.5 and 5-8.5 scenarios. The CMIP6 historical data is also compared locally at seven measurement stations. Two future projections are considered for the near-term (2031–2060) and mid-term (2071–2100), focusing on inter- and intra-annual variability and extreme wind behaviour. The results indicate an overall decrease in mean wind speed across the study area, with more pronounced reductions under SSP5-8.5 and during the mid-term period. In terms of seasonality, reductions are more pronounced during winter and summer than in the transitional seasons. According to the extreme value analysis based on the generalised extreme value distribution, general declines in extreme values are detected at selected critical locations, especially at wind speeds with large return periods. Full article

Background: The global aging population faces an increasing prevalence of type 2 diabetes mellitus (T2DM), often complicated by frailty, cognitive decline, and impaired manual dexterity. These factors make glucose self-monitoring particularly challenging. Minimally invasive glucose monitoring methods, particularly continuous glucose monitoring (CGM) as well as emerging non-invasive glucose monitoring technologies offer potential solutions, but remain insufficiently evaluated in older adults. Objective: To systematically review and synthesize available evidence on the advantages of continuous glucose monitoring (CGM) and non-invasive glucose monitoring methods in older adults aged ≥65 years, focusing on clinical efficacy, usability, adherence, and existing knowledge gaps. Methods: A systematic literature search was conducted across PubMed, Scopus, and Web of Science, including studies from 2020 to 2025. Eligible studies included participants aged ≥65 years and evaluated the clinical performance of CGM or other minimally invasive or non-invasive glucose monitoring technologies. The PRISMA framework guided screening and selection. Risk of bias was assessed using RoB 2 and ROBINS-I tools. Due to substantial heterogeneity among study designs and reported outcomes, a narrative synthesis approach was adopted. Results: A total of 426 records were identified, of which 13 met the predefined eligibility criteria after full-text screening. After risk of bias assessment, one study was excluded, resulting in 12 studies included in the final synthesis. No eligible studies evaluating completely non-invasive glucose monitoring technologies were identified, highlighting a significant research gap in this area specifically for older adults. CGM was associated with improved glycemic control, reduced hypoglycemia, and increased time in range among older adults. Usability was generally high, particularly with newer, user-friendly devices. Conclusions: CGM is associated with improved glycemic outcomes and favorable usability in adults aged ≥65 years. However, a significant gap exists in research on non-invasive glucose monitoring technologies in this population. Future studies should address the accuracy, feasibility, and usability of non-invasive glucose monitoring devices, while accounting for the physiological and behavioral complexities associated with aging. Full article

The Technology Acceptance Model (TAM) has been widely used to explain e-banking and digital technology adoption. Existing literature supports the robustness of its core relationships, but the magnitude of the effects varies considerably across studies, raising questions about their stability and generalizability in new contexts. Existing meta-analysis studies focus primarily on pooled effect sizes, providing limited insight into the temporal stability of relationships, their sensitivity to individual studies, and the extent to which observed heterogeneity reflects contextual variation. This study contributes by reinterpreting heterogeneity not as a problem to be reduced, but as a feature that defines the limits of generalizability. We advance the TAM literature by moving beyond average effects and rethinking empirical evidence through the joint lens of robustness, stability, and dispersion. We conduct a random-effects meta-analysis on 44 effect sizes (correlation coefficients) coming from 43 research papers indexed in Web of Science and Scopus. In addition to pooled correlations, the analysis employed cumulative meta-analysis, leave-one-out influence diagnostics, prediction intervals, and publication bias assessments to evaluate the evolution, consistency, and variability of TAM relationships across contexts. The findings show that core TAM relationships are consistently positive and stable at the aggregate level yet display substantial variation across empirical settings. While some relationships remain robust across contexts, others exhibit prediction intervals that include zero, indicating that their strength and even direction may depend on contextual conditions. As prior TAM meta-analyses have not systematically incorporated prediction intervals, this study provides new evidence to the extent to which TAM relationships generalize beyond average effects. The results further show that although TAM offers a reliable structural framework, interventions and policies based on its core relationships must be context-sensitive, because relying on average effects alone may lead to ineffective or inconsistent adoption outcomes. Full article

Ocean surface currents (OSCs) are central to upper ocean dynamics and air–sea exchange, yet their retrieval from spaceborne synthetic aperture radar (SAR) is limited by wave-induced bias (WB). WB arises from the inherent motion of the scattering facets and from long-wave hydrodynamic and tilt modulations, and is therefore jointly controlled by sea state and radar viewing geometry. This study develops an observation geometry optimization framework. Five years of hourly ERA5 wind and wave reanalysis data over the Kuroshio are used as a representative ensemble of sea states to drive the KaDOP model, and an exhaustive grid search over line-of-sight (LOS) azimuth (0–360°) and incidence angle (20–60°) is performed to identify, for each location and season, the viewing geometry that minimizes the time-mean WB. These local optima are then summarized as mission-level metrics, including the minimum achievable WB, the coverage meeting prescribed WB thresholds, and the spatial coherence of the preferred LOS azimuth and incidence angle. Finally, the theoretical minima are compared with the fixed left-looking geometry of the Luojia-2 (LJ-2) satellite along a 213 km × 6 km observation corridor and with Gaofen-3 (GF-3) viewing geometries at four representative locations in the Kuroshio. Across these validation cases, the optimized geometry reduces mean absolute WB by about 20–60% for LJ-2 and 20–80% for GF-3, providing quantitative constraints for future SAR mission design targeting OSCs. Full article

Owing to the fact that the conventional Temperature Drift Errors (TDEs) precise estimation model of Capacitive MEMS-Gyros (CMG) has incomplete Temperature Correlated Quantities (TCQ) and an imperfect parameter identification method to reduce bias stability, a BO-LSTM-based TDE precise estimation model using thermal-induced physical characteristics variation analysis is proposed. By analyzing microstructural deformation in CMG- and Si-based materials’ stiffness deterioration caused by thermal-induced physical characteristics variation, complete TCQ are traced, including ambient temperature variation ∆T and its square root ∆T^1/2 plus its higher orders (∆T², ∆T³, ∆T⁴), a modified TDE precise estimation model is formed. Long Short-Term Memory (LSTM) is applied to identify the modified model’s parameters owing to the typical time series characteristics of TDE and TCQ. In addition, Bayesian Optimization (BO) is introduced in LSTM to show a good guide for LSTM’s optimal hyperparameters. The modified model is implemented with BO-LSTM and compared with the conventional model based on Radial Basis Function Neural Network (RBFNN) in bias stability. The experimental results show that the modified model can more accurately estimate the TDE of CMG in a timely and improves its bias stability by 20%, which decouples the temperature dependence of Si-based material significantly and enhances the environmental adaptability of CMG in complex conditions remarkably. Full article

In this study, a simulation-based investigation of the variations of the bit error rate (BER) and the maximum quality factor are presented for short- (S-), conventional- (C-), and long- (L-) band wavelengths in a photonic ultra-wideband (UWB) circuit using a semiconductor optical amplifier (SOA) with different bias currents and a bandpass filter (BPF). Gaussian quadruplet UWB pulses are generated at the S-, C-, and L-band wavelengths, which are commonly used in fiber transmission lines. An analysis of the temporal and spectral features of the generated pulses is carried out. The highest maximum quality factor and the lowest minimum BER are obtained in the C-band at an SOA bias current of 150 mA. This study simultaneously investigates both UWB pulse generation and transmission performance. The proposed circuit has a simple design and high applicability, as it employs a SOA, a Gaussian optical filter, a low-pass filter (LPF) and a single BPF. Full article

The immunosuppression inherent to chronic lymphocytic leukemia (CLL) is the most frequent cause of morbidity and mortality associated with this pathology. Both the innate and adaptive immune responses exhibit marked functional alterations, with a bias towards a tolerant environment that favors the spread of the disease. This condition is reflected in increased risk of infections, immune-mediated cytopenias, and associated second malignancies. Knowledge of these alterations, both in the molecular pathways that modulate T cell activity in CLL (the T lymphocyte cytotoxic antigen-4 (CTLA-4) axis and programmed cell death 1 (PD-1)) and at the T cell immunoreceptor level, could be of interest as therapeutic targets in CLL. In this review, we will analyze the main consequences of this dysfunction and its management strategies. Full article

Accurate distinction between non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC) remains the key local staging problem in bladder cancer because treatment intensity, timing of radical therapy, and suitability for bladder-preserving strategies all depend on it. Multiparametric magnetic resonance imaging (mpMRI) and the Vesical Imaging-Reporting and Data System (VI-RADS) now provide a standardized imaging framework for local staging and increasingly support MRI-first clinical pathways. Artificial intelligence (AI) has emerged as an additional decision-support layer, but the evidence base remains methodologically uneven. In this structured narrative review, we synthesized peer-reviewed literature from January 2020 to March 2026, while retaining foundational VI-RADS studies from 2018 to 2019, and prioritized guideline documents, meta-analyses, prospective cohorts, multicenter and externally validated AI studies, response-assessment studies, and papers addressing implementation and reporting quality. Current evidence shows that radiomics and deep learning models can achieve high discrimination for MIBC detection on MRI, and that the most plausible incremental value of AI lies in equivocal VI-RADS lesions, reader support outside high-volume expert settings, and multimodal risk stratification. However, most studies remain retrospective, highly selected, segmentation-dependent, and vulnerable to reference-standard bias, domain shift, and poor calibration. This review therefore emphasizes several translational issues that are often underreported: lesion-level versus patient-level inference, the distortive effect of TURBT-based labels, the need to evaluate false-negative consequences in VI-RADS 3 tumors, and the distinction between diagnostic support and broader pathway redesign. We also discuss response assessment, nacVI-RADS, segmentation automation, multicenter and federated infrastructure, workflow ownership, and the limits of imaging-only models in a biologically heterogeneous disease. The most credible near-term role of AI is not autonomous diagnosis, but augmentation of standardized mpMRI and VI-RADS within multidisciplinary care. Future progress will depend on prospective utility studies, site-held-out validation, transparent reporting, and the integration of imaging with molecular and cellular heterogeneity through radiogenomic and multi-omics approaches. Full article

To address the core challenges in air-to-ground target re-identification (ReID), including network focus on invalid background information, poor adaptability to nonlinear feature distribution, and insufficient cross-domain generalization, this paper proposes a novel air-to-ground ReID framework based on Neighborhood Feature Centralization Attention (NFCA). On the basis of Coordinate Attention, the framework introduces a parameter-free Neighborhood Feature Centralization mechanism to build a lightweight attention module, which enhances cross-feature semantic interaction and suppresses background noise while retaining precise position encoding. It achieves end-to-end direct optimization of sample pair similarity through binary cross-entropy loss, eliminating the proxy task bias of traditional classification loss and adapting to the nonlinear structure of feature space. A multi-source data-driven training strategy is constructed by fusing ReID datasets and general classification datasets, which expands the coverage of feature space and narrows the distribution gap between training data and real air-to-ground scenarios without additional manual annotation. Experiments show that the proposed method achieves leading mAP values on the self-developed UAV air-to-ground dataset JC-1, the public person ReID dataset Market-1501, and the public vehicle ReID dataset VehicleID. Sufficient statistical validation, ablation experiments and cross-domain tests verify the advancement, reliability and generalization of the proposed method in complex air-to-ground scenarios. Full article

Kang Youwei (1858–1927) reimagined Confucius as the founding religious leader of Confucianism, a conceptual framework underpinning his entire ideological system of Confucian thought. Yet existing scholarship has largely overlooked systematic analysis of this theoretical reconstruction. Influenced by the impact–response paradigm, many studies have also neglected Kang’s core intention to pursue cross-civilizational dialogue and establish a universalist Confucianism through such interpretive innovation. Faced with the late-Qing predicament of the imbalance between a dominant Western world and a weakened China, Kang thoroughly redefined Confucius by shifting his image from a sage who transmitted rather than created ancient wisdom to a religious authority who reformed institutions through classical precedents. This paper argues that Kang’s reinterpretation was neither a simplistic religious adaptation nor a conservative defence of traditional culture. His fundamental aim was to correct Western-centric bias, facilitate equal Sino-Western civilizational dialogue, critique inherent structural dilemmas of modern Western civilization, and propose the Confucian Way as a viable solution to these deep-rooted crises. Full article

Background: With the increasing demand for non-surgical facial rejuvenation, injectable collagen biostimulators such as poly-L-lactic acid (PLLA), calcium hydroxyapatite (CaHA), polycaprolactone (PCL), poly-D,L-lactic acid (PDLLA) and powdered polydioxanone (PPDO) have become widely used by facial aesthetic practitioners. These agents stimulate neocollagenesis, providing gradual improvement in skin firmness, elasticity and facial contour with long-lasting results. While manufacturers emphasize the efficacy and favorable safety profile of these products, adverse events such as nodules, edema, inflammatory reactions and, in rare cases, granulomas have been reported. To date, no comprehensive systematic review has evaluated the proportion and nature of adverse effects associated with all major collagen biostimulators in facial aesthetic procedures. This study aims to synthesize current evidence on the proportion of adverse events linked to injectable collagen biostimulators. Methods: The systematic review will include clinical studies involving adults undergoing facial aesthetic procedures with PLLA, PDLLA, CaHA, PCL and PPDO that report adverse events during or after treatment. The search will be conducted in six main databases: CENTRAL, EMBASE, LILACS, PubMed, SCOPUS and Web of Science. No restrictions will be applied regarding language or publication date. The screening process will occur in two phases: first, two independent reviewers will assess titles and abstracts against the eligibility criteria; second, the same reviewers will conduct full-text evaluations. Data will be synthesized narratively, with a meta-analysis of proportions performed if appropriate. Additionally, sample characteristics, treatment protocols, study design and main findings will be reported. The risk of bias will be assessed independently by two reviewers using appropriate tools, based on the study design, with the support of artificial intelligence. PROSPERO registration number: CRD420251062785. Full article

Background: Stretching exercises are strongly recommended as part of exercise training programs; however, their effects on blood pressure (BP) and other related cardiovascular parameters in adult individuals with elevated BP (pre-hypertension) or hypertension remain unclear. Methods: A systematic search was conducted in PubMed and databases accessed via the EBSCO platform up to 30 September 2025, following the PRISMA guidelines. An additional search of Scopus was performed on 8 April 2026. Studies eligible for inclusion were randomized controlled trials, randomized crossover trials, non-randomized clinical trials and single-arm trials investigating stretching interventions in adults with pre-hypertension and or hypertension. Risk of bias assessment was performed using RoB 2 for randomized trials and ROBINS-I for the non-randomized trials. A random-effect meta-analysis was performed when at least two studies reported sufficiently comparable BP outcomes. The quantitative synthesis was considered exploratory. Results: Eleven records published between 2014 and 2025 met the eligibility criteria and were included. All protocols used static stretching, although only a small number were clearly described as active stretching. The results were heterogeneous across the design, duration of intervention and outcomes. Chronic interventions more often reported favorable changes in indices of arterial stiffness, whereas acute interventions demonstrated more variable immediate BP responses. In the exploratory meta-analysis, the pooled estimate suggested a reduction in systolic blood pressure (SBP) in favor of stretching; however, this effect did not reach statistical significance (mean difference (MD) = −5.39 mmHg, 95% confidence interval (CI): −11.32 to 0.53; I² = 0%). For diastolic blood pressure (DBP), the pooled estimate favored stretching and reached statistical significance (MD = −3.93 mmHg, 95% CI: −7.25 to −0.60; I² = 0%). In sensitivity analyses including a third study, the pooled effects remained in favor of stretching for systolic BP (MD = −6.6 mmHg, 95% CI: −12.2 to −1.0; I² = 56%) and diastolic BP (MD = −5.4 mmHg, 95% CI: −7.1 to −3.7; I² = 8%). These pooled estimates should be interpreted with caution due to the small number of studies, heterogeneity in study design and participant characteristics, and overall limitations in methodological quality. Secondary findings suggested possible improvements in selected vascular parameters, including brachial–ankle pulse wave velocity, augmentation index, and cardio–ankle vascular index, whereas acute responses were more variable and protocol-dependent. Overall, the level of evidence was limited, with most randomized trials judged as having some concerns and non-randomized studies judged as having a critical risk of bias. Conclusions: Stretching interventions may improve BP and selected vascular parameters in adults with pre-hypertension and hypertension and may represent a practical adjunct within the non-pharmacological management of BP. However, the current evidence is limited by methodological heterogeneity, risk of bias, and the small number of studies available for quantitative synthesis. Therefore, the pooled findings should be considered exploratory and hypothesis-generating rather than definitive. Further high-quality randomized controlled trials are required to determine the optimal type, dose, and long-term clinical relevance of stretching interventions in this population. Full article

In sandstorm environments, a large number of suspended particles in the air absorb and scatter light, causing strong color bias, low contrast, and blurred details in images. These degradations reduce the reliability of computer vision applications in surveillance systems, intelligent transportation systems, unmanned aerial vehicle monitoring, and outdoor autonomous driving systems. A complete sandstorm image enhancement method was developed in this study by combining sky detection, color correction, contrast enhancement, and adaptive dark channel prior (ADCP) dehazing. The Lab color space was used to correct the color bias. The L channel was enhanced using normalized gamma correction and contrast-limited adaptive histogram equalization to improve brightness and contrast. Then, the sky region is detected to avoid over-processing, preserving the natural appearance of the sky region. Finally, ADCP is applied to non-sky regions for further dehazing. Experiments show that the proposed method provides better subjective and objective performance compared to other algorithms. Full article