Search Results (574)

Longline fisheries harvest high-value pelagic and demersal species but remain labor-intensive and require reliable monitoring of catches, bycatch, and fishing effort. This technology-oriented narrative review synthesizes 71 sources on mechanization, automation, electronic monitoring (EM), artificial intelligence (AI), and prospective smart fisheries. Evidence is distinguished by source type, application setting, and operational maturity. Mechanized hauling, baiting, and integrated autoline systems support fishing operations, whereas EM provides records for catch verification and management. Computer vision studies demonstrate catch-event detection, species classification, and selected compliance-monitoring tasks, but reported performance remains specific to the datasets and validation conditions. Commercial availability, experimental performance, and conceptual feasibility are therefore assessed separately. Edge AI, integrated intelligent reporting, digital twins, and generative AI remain emerging or prospective for longline fisheries in the reviewed evidence. The synthesis identifies camera visibility, cross-vessel generalization, rare-species detection, interoperability, and data governance as key constraints. A practical development pathway combines established equipment and EM with externally validated AI, expert review, and feedback to subsequent vessel operations. Increased fishing efficiency alone does not establish sustainability; environmental and management outcomes require separate evaluation. Full article

Background/Objectives: Cervical cancer prevention relies on cytology, high-risk human papillomavirus (HR-HPV) testing, and histological confirmation, but these modalities may disagree. We quantified test concordance and determinants of CIN2+ in a specialized gynecologic referral cohort with very low HPV vaccination coverage. Methods: We retrospectively analyzed 172 women evaluated for suspected or established cervical pathology between 2022 and 2024. Of these, 160 had a definitive HPV result and 130 had histology; histology-based performance analyses used complete cases and are conditional on biopsy verification. Cytology was categorized by the Bethesda system, HR-HPV genotypes were recorded, and histology was dichotomized as CIN2+ versus ≤CIN1. Results: Among 160 women with definitive HPV results, 109 (68.1%) were HPV-positive. HPV-positive women were younger (median 33 vs. 43 years, p < 0.001) and, among biopsied women, more frequently had CIN2+ (70.2% vs. 22.2%, p < 0.001). Cyto-histological agreement was fair (65.0%, κ = 0.32). HPV16/18 was strongly associated with CIN2+ (OR 6.85, 95% CI 2.54–18.49). Within histologically verified complete cases, HPV DNA positivity had the highest apparent sensitivity (91.7%), whereas HPV16/18 had the highest PPV (86.0%). Conclusions: Cytology and HR-HPV testing provide complementary information in this referral-enriched cohort. Because disease prevalence was high and biopsy verification was selective, predictive values and apparent diagnostic-performance estimates should not be extrapolated directly to population screening; external validation in primary-screening populations is required. Full article

The use of artificial intelligence (AI) and machine learning (ML) models has increased significantly in recent years, with rapid advances across domains such as natural language processing, computer vision, code generation, and scientific computing. Alongside this growth, a diverse ecosystem of models and services has emerged, raising important questions about efficient and fair mechanisms for accessing and paying for inference. In particular, there is an ongoing debate between subscription-based access and per-request pricing models, the latter becoming especially relevant in the context of autonomous AI agents that dynamically consume external services. One of the emerging open protocols for enabling per-request payments is x402, which uses Hypertext Transfer Protocol (HTTP)-native payment flows. The core payment protocol does not let clients verify that inference was performed using the claimed model. In this paper, we propose a framework for verifiable paid inference that integrates x402-based micropayments with verifiable ML techniques. The framework combines a payment authorization bound to model and input commitments with a common escrow design for synchronous and asynchronous verification. Payment is released after a valid proof or after an optimistic challenge process accepts the result under its stated assumptions. The prototype demonstrates the functional feasibility of this approach, together with measured gas costs and proving times. The measurements indicate that asynchronous verification is more suitable for larger workloads and higher request volumes, whereas full-model proving is practical primarily for smaller models. Full article

Mountain-greenway planning requires approaches that account for heterogeneous user needs, trail safety, experiential quality, and limited maintenance capacity. Taking the Shenzi Village–Longshan Village section of Wugong Mountain as a case, this study develops a perceived-value-informed, segment-level, scenario-based framework for assessing mountain greenway suitability. COROS PACE 3 FIT trajectories, field observations, and researcher-based environmental interpretation were used to construct eight indicators for ten segments. Entropy weights characterize within-sample variation, while the Potential Suitability Value (PSV) integrates environmental conditions with a priori utility rules for low-, moderate-, and high-activity scenarios. Ranking sensitivity was examined through utility-function sampling, fixed-endpoint comparisons, equal-weight comparisons, leave-one-segment-out reweighting, weight perturbation, and indicator-deletion analysis. Shade, exposure, movement difficulty, and cumulative ascent exhibited relatively high differentiation within the present sample. S10 ranked among the top three under all three scenarios in both the baseline entropy-weight and leave-one-segment-out schemes, but fell to fourth under the high-activity scenario with equal weights. The low-activity rankings of S5 and S8 were jointly affected by the span and internal spacing of the utility functions, whereas the high-activity advantage of S6 depended structurally on the treatment of risk indicators. By identifying segments with inadequate environmental support, elevated safety pressure, or parameter-sensitive rankings, the framework can inform targeted field verification, maintenance prioritization, and risk communication. It is a scenario-based suitability and planning-support tool, not an assessment of sustainability outcomes. Ecological disturbance, environmental carrying capacity, community benefits, social equity, long-term maintenance costs, resource efficiency, and actual user perceptions were not measured and require separate indicators and external validation. Full article

Background/Objectives: Computer-assisted posturography is utilized to quantify human postural control, yet existing dedicated systems frequently present physical constraints, such as limited sensing surfaces and rigid hardware barriers. In this study, a methodological framework is presented and a structural arrangement (OpenBalance) is designed to implement balance assessment accommodating a wider anthropometric range through zone-based foot placement using a decentralized load cell array, with preliminary implementation details provided in a repository. Methods: The configuration comprises an array of 16 discrete uniaxial vertical-force load cells embedded within four mechanically decoupled sub-platforms to minimize mechanical cross-talk. Biomechanical moment equations were implemented in a custom Python pipeline to compute a localized Center of Pressure (COP) for each sub-platform independently. A vector-based data fusion algorithm maps these local coordinates into a unified global coordinate system. The system evaluation incorporated baseline signal-to-noise ratio (SNR) analysis, mechanical crosstalk testing, and digital low-pass filtering. Results: A proof-of-concept evaluation using empirical data confirmed that the cascading coordinate model produces a continuous global COP trajectory and quadrant-specific load distributions. The platform dimensions (355 × 460 mm) and sensor topography geometrically accommodate natural external foot rotation (incorporating a 10° toe-out angle projection) and foot lengths corresponding to EU shoe sizes up to 55. Mechanical crosstalk between adjacent sub-platforms remained minimal (<1% of applied load). Conclusions: The OpenBalance framework confirms the technical feasibility of deriving a continuous global COP from a decentralized array of distributed load cells. While baseline component specifications and static verifications are established, comprehensive dynamic cross-validation against reference standards remains a necessary next step for the future development of open hardware that can be produced using 3D printing. Full article

Electric vehicle charging equipment now links vehicles, payment and roaming services, cloud platforms, and the grid, but its cybersecurity requirements are fragmented across standards. This study combines a cross-standard gap analysis with a traceable conformity pre-assessment framework for a local large language model; the framework is a design-validation proof of concept, not a production-ready conformity-assessment system. Using EN 18031 as the reference axis, IEC 62443-4-2, ETSI EN 303 645, and NIST IR 8473 requirements were mapped on a five-level correspondence scale by a single researcher and consolidated into integrated requirements with applicability, evidence, and acceptance rules. A clause-aware retrieval-augmented generation pipeline (semantic retrieval only) was compared with no-retrieval and fixed-length baselines on 80 synthetic design-gold cases, reported on the 44 holdout cases not used for tuning. Better retrieval did not ensure trustworthy conformity reasoning: retrieved clause text in the judgment prompt lowered verdict accuracy (0.89 without, 0.77 with), the most accurate configuration (0.93) used retrieval only for citations, and a rule-based classifier with no model or corpus reached 0.66 on the same cases. Citation extraction, citation verification, and human review were each necessary. Extraction guarantees citation existence, but exact-gold-clause agreement remained at most 0.38. External validity on real manufacturer documents is untested; the framework complements rather than replaces accredited certification. Full article

Background/Objectives: Multimodal artificial intelligence (AI) is increasingly used in radiology to combine medical images with radiology reports, clinical narratives, structured health records, laboratory measurements, and other patient data. These systems may support more context-aware interpretation, reporting, and clinical decision-making than unimodal approaches. This narrative review examines recent technical and clinical advances in multimodal radiology AI and identifies barriers to responsible implementation. Methods: Relevant biomedical and technical literature was identified through targeted searches of PubMed, IEEE Xplore, ACM Digital Library, Web of Science, Scopus, arXiv, ScienceDirect, SpringerLink, and Google Scholar. A documented screening process identified 111 reviewed publications from 1154 records. Study selection and data extraction were conducted by one reviewer without independent verification. Original research, reviews, and commentaries, including selected preprints, were synthesized thematically. Results: The field has progressed from early and late feature fusion toward cross-modal attention, contrastive image–text pretraining, vision–language models, multimodal large language models, and general-purpose foundation models. Applications include diagnostic classification, prognostic modelling, image–text retrieval, visual question answering, clinical decision support, and automated radiology report generation. Despite promising technical results, comparison across studies remains difficult because of heterogeneous datasets, tasks, metrics, and validation designs. Clinical translation is further limited by scarce external and prospective validation, uncertain interpretability, hallucination and omission risks, privacy and fairness concerns, and limited real-world workflow evaluation. Conclusions: Multimodal AI may enable more clinically informed radiological interpretation and reporting, but progress in benchmark performance has outpaced evidence of safety, generalizability, and clinical utility. Future research should prioritize multicentre evaluation, clinically meaningful metrics, transparent reporting, robust safety assessment, and workflow-centred implementation. Full article

Background: Online postgraduate (PG) public health education must develop higher-order competence and readiness for practice. Generative artificial intelligence (genAI) may enhance feedback, simulation and scaffolding, but may also substitute for learner reasoning. This review examined pedagogical conditions supporting higher-order learning and how AI interacts with them to influence competence and real-world readiness in online PG public health education. Methods: Following JBI guidance and PRISMA-ScR, PubMed and Scopus were searched for literature published from 2010 to 2026, with the final search conducted on 20 August 2026. Eligible sources addressed higher-order learning, authentic learning, competence or professional readiness in PG public health education, or provided transferable evidence from health professions and higher education settings. Of 5777 records identified, 253 duplicates were removed and 5524 screened. Data were charted using a structured evidence matrix and stratified into direct, transferable, synthesis, and contextual evidence layers. Results: Of 401 records undergoing detailed eligibility and analytic-role assessment, 338 were retained. Core evidence comprised 92 studies: 27 direct and 65 transferable; 32 syntheses provided triangulation. AI featured in 11.1% of direct versus 87.7% of transferable studies. Applied competence/skills occurred in 59.8%, synthesis/integration in 44.6%, critical thinking in 26.1%, and real-world cases/scenarios in 51.1%. Practice/workplace transfer was more prominent in direct than in transferable evidence (66.7% vs. 13.8%). Direct AI evidence in PG public health was limited, and heterogeneity across studies constrained direct comparison and precluded quantitative pooling. Conclusions: Higher-order learning depended more on pedagogical architecture than modality or AI adoption. AI may amplify cognition when preserving learner reasoning, verification, and judgement, but may substitute for cognition when generated outputs replace these processes. No external funding was received for this review. Full article

Lightweight UAVs increasingly operate in constrained environments, where even minor collisions can damage exposed propellers, arms, sensors, or embedded components. Designing protective cages for such platforms requires more than adding an external shell: the structure must preserve propeller clearance, fit the chassis, remain lightweight, avoid obstructing sensors and cameras, and remain feasible to prototype. This paper introduces PICLIN (Pipeline for Incremental Constraint-Linked InnovatioN), a graph-based design-to-prototyping workflow for UAV protective cages. PICLIN starts from abstract topological representations and progressively constrains them through geometric deployment, curved-surface embedding, numerical refinement, drone-integrated verification, and physical realization. The approach is demonstrated through a UAV case study leading from graph exploration to assembled protective prototypes. Results show that PICLIN generates coherent cage configurations while preserving early-stage design freedom and reducing the solution space toward manufacturable structures. For the representative Optimal 9 prototype, the cage mass was 151 g, excluding the current 28 g aluminium UAV interfaces, with an overall diameter of 485 mm and a design clearance of 50 mm between the propellers and the protective cage. A preliminary 1 m drop test of the 1.186 kg cage–chassis assembly, corresponding to an impact energy of approximately 11.6 J, produced a peak acceleration of approximately 35 g over an impact interval of about 15 ms. These measurements are reported as proof-of-feasibility results rather than as a complete mechanical validation. The contribution is not a single optimized cage geometry, but a reusable workflow linking abstract graph modeling, constraint integration, UAV-specific verification, and prototype-oriented realization. Full article

Clinical artificial intelligence (AI) has demonstrated performance across diagnostic, predictive, monitoring, and decision-support tasks, yet hospital adoption depends on whether that capability can produce sustainable value under local conditions. Informed by structured searches of PubMed/MEDLINE and Scopus, supplemented by Google Scholar and citation tracking (2015–July 2026), this narrative review synthesizes evidence on clinical AI methods, applications, implementation, evaluation, and sustainability in hospitals. Evidence remains uneven: image-based applications have substantial evidence for diagnostic performance, whereas prospective clinical utility, workflow effects, economic consequences, and organizational sustainability are evaluated less consistently. Established approaches—including efficacy hierarchies, the Radiology AI Deployment and Assessment Rubric (RADAR), hospital-based health technology assessment, implementation science, and AI governance frameworks—offer perspectives on performance, transferability, implementation, and value. Building on these traditions, this review proposes institutional efficacy as the warranted and revisable hospital-level judgment that an AI-enabled service configuration can generate and sustain patient-centered value under local priorities, constraints, and adaptive capacity. The review further proposes that deeper organizational integration increases the proportion of realized-value evidence requiring local generation or verification. Two additional properties—context and data dependence, and the consequence profile—separately shape that burden. External evidence remains essential, but hospital-specific appraisal is necessary to determine whether value can be realized and sustained. Full article

Background: Counterfeit goods realize illicit value only when moved, mainly by logistics intermediaries rather than counterfeiters, so how the law treats these intermediaries determines where enforcement can bite. This review traces the policies, laws, literature, and data relevant to China’s anti-counterfeiting landscape from 1979 to 2026. Methods: A narrative review combined a PRISMA 2020-guided search of Scopus and Web of Science (108 peer-reviewed publications) with 14 primary legal instruments and 4 types of institutional grey literature. Institutional theory was used to derive four expectations. Results: Four regulatory phases were identified, across which duties migrated toward logistics chokepoints: criminal accomplice liability (2004), civil facilitation liability, customs detention, platform liability, parcel-level identity, inspection and screening duties, franchisor network supervision (2024), and independent liability for assisting confusion (2025). Expectations of external origin, chokepoint migration, and ceremonial compliance were supported; capacity-calibrated decoupling awaits verification. Obligations were synthesized into an actor-by-actor matrix. Official data show that only about 3% of administrative cases reach criminal referral, with no breakdown by intermediary type. Conclusions: China has converted logistics intermediaries from neutral conduits into liability-bearing compliance nodes under a know-your-consignor regime calibrated to large integrated operators, but this regime still relies primarily on coercive measures. Full article

Bioregenerative Life Support Systems (BLSS) are considered a key technology for long-duration human space exploration due to resource transport limitations. However, they require a systematic characterization and qualification before they can be integrated into large-scale demonstrators or analog facilities. Reproducible ground-based testing environments are therefore needed to maintain controlled boundary conditions while allowing payload-induced gas exchange to be discerned from external disturbances. This study presents a modular characterization environment developed for the qualification of BLSS payloads and technologies, with particular emphasis on cabin atmosphere gas exchange. The platform provides independent control of O₂, CO₂ and N₂ contents, thermal and humidity management and a distributed sensing architecture with database-integrated data logging to support long-duration operation. System performance was verified through long-duration experiments addressing cabin atmosphere stability, leakage-induced concentration drift, gas-composition control, disturbance response and thermal management. Defined O₂ and CO₂ concentration offsets were maintained over extended operating periods ranging from days to weeks with passive drift low enough to enable quantitative payload gas-exchange characterization. Stress testing under simulated O₂ consumption demonstrated robust controller performance under elevated disturbance loads. Thermal verification confirmed preservation of payload-specific temperature control, controlled condensate recovery and humidity management. These results demonstrate that the platform provides stable, configurable and reproducible boundary conditions for quantitative testing of BLSS payloads. By bridging laboratory-scale development and integrated BLSS facilities, the characterization environment enables technology qualification for full-scale implementation. Full article

Efficient deployment of neural-network detectors on field-programmable gate array (FPGA) accelerators depends not only on model complexity but also on compiler-visible graph structure. On deep learning processing unit (DPU) platforms, unsupported operators can fragment execution between accelerator and host execution domains. We present C2fDeploy, a training-free rewrite for the YOLOv8 C2f block that moves channel splitting from the activation graph to an offline partition of the trained projection and batch-normalization parameters. The transformed block preserves the 32-bit floating-point (FP32) function without retraining or additional parameters. On the GRAZPEDWRI-DX fracture-detection task, the original and rewritten graphs produced identical FP32 test metrics and comparable 8-bit integer (INT8) accuracy. Direct tensor-level FP32 comparison at the outputs of all eight rewritten C2f blocks yielded an aggregate mean absolute error of $9.73\times {10}^{-8}$ and a relative $L_2$ error of $1.97\times {10}^{-7}$, providing numerical verification beyond detection-level metrics. Compilation for the Kria KV260 consolidated nine DPU subgraphs into one and removed the C2f-related host-side slicing operations. In a same-checkpoint whole-XModel benchmark, C2fDeploy improved whole-XModel graph execution throughput by $95.2\times$ and reduced energy per execution on the 5 V system-on-module (SOM) rail by 98.2%. Direct runtime profiling further showed that DPU compute-unit busy-time utilization increased from 0.36% to 90.75%, while system-wide CPU utilization decreased by 89.5%. Aggregate APM-observed external-memory bandwidth increased from 40.20 to 3250.56 MB/s as accelerator execution became more continuous, whereas normalized APM-observed traffic decreased by 14.8% per graph execution. These results show that compiler-aware graph rewriting can remove deployment bottlenecks without changing the trained detector. Full article

Cardiac autonomic recovery after exercise is conventionally summarized as a single nocturnal heart rate variability (HRV) average, treating sleep as a passive backdrop rather than an active determinant of recovery. Three distinct levels of inference are involved and are kept separate throughout: an observed association between sleep stage and autonomic indices; causal evidence that specific experimental manipulations of slow-wave sleep alter autonomic or cardiovascular parameters; and the substantially more tentative, not-yet-demonstrated claim that slow-wave sleep plays a causal role in recovery, specifically after exercise. A fragmented but convergent literature indicates that autonomic state tracks sleep stage in real time, with HRV indices of cardiac-vagal modulation concentrated in slow-wave sleep (SWS) and attenuated during REM sleep, and that the SWS–autonomic link itself has been demonstrated experimentally, not only observed: enhancing slow-wave activity increases HRV-derived cardiac-vagal modulation, and sleep restriction directly disrupts the nocturnal autonomic state. Whether this causal pathway produces post-exercise autonomic recovery remains to be demonstrated. Exercise has been shown to modulate SWS-phase HRV, and post-exercise vagal reactivation has been reported to be depressed on nights following intense exercise in a manner localized to the SWS stage, though this rests on cardiac-derived rather than EEG-confirmed sleep staging. However, no study has manipulated SWS after exercise and tracked the resulting recovery trajectory, a gap that this review names explicitly (as Prediction 1, P1) as the framework’s most direct causal test. This review synthesizes the literature into a systems physiology framework in which SWS is hypothesized to function as a plausible moderator of post-exercise cardiac autonomic recovery, not yet as a demonstrated determinant of it, integrating brainstem circuitry and neurotransmitter systems, causal SWS manipulations (acoustic, pharmacological, deprivation-based), exercise dose-response, and boundary conditions (age, sex, training status, sleep and cardiovascular disorders) under which the coupling is preserved, attenuated, or absent. The effect is graded and window-specific rather than uniform—comparatively strongest in the early post-exercise reactivation phase and during nocturnal SWS, not established across the full multi-hour recovery curve—and is translated into falsifiable predictions and candidate study designs, including a critical appraisal of wearable sleep-tracking validity. Two evidentiary gaps are addressed transparently: the human circuit-level mechanism remains largely inferred from rodent studies, and independent citation-network verification is not feasible for all sources. These findings support slow-wave sleep as a plausible, mechanistically motivated candidate moderator of post-exercise systemic autonomic and cardiovascular physiology, rather than an established determinant of it or a variable external to it. Full article

The increase in the volume of organic household waste and the transition to a circular economy require the development of reliable forecasting tools capable of ensuring effective planning of waste recycling and bioenergy production systems. Despite significant advances in machine learning methods, most current research focuses on improving forecasting accuracy, while paying insufficient attention to model interpretation and the use of the results to support management decisions. The aim of this study is to develop and validate, using empirical municipal data, an explainable machine learning framework for forecasting the generation of organic household waste to support decision-making regarding the development of biogas and biomethane systems. The proposed framework combines data preprocessing, feature engineering, ensemble machine learning algorithms, prediction quality assessment, Explainable Artificial Intelligence (XAI) methods, and bioenergy potential assessment into a unified decision support system. The study was conducted using an empirical dataset comprising 10,800 observations for 20 local communities, obtained from the monitoring and accounting records of LKP “Green City” (Lviv, Ukraine) and covering the period from 1 January 2024 to 23 June 2025. The dataset includes demographic, socioeconomic, territorial, tourism-related, natural and climatic, infrastructural, logistical, organizational and economic, and temporal characteristics. To forecast the normalized index of organic household waste generation, we developed and compared Random Forest, Extra Trees, Gradient Boosting, XGBoost, LightGBM, and CatBoost models. A Linear Regression baseline fitted to the overlapping index-related predictors achieved MAE = 0.02339, RMSE = 0.02886, and R² = 0.96412 on the chronological holdout set, indicating that a substantial part of the constructed target is linearly reconstructible. Among the six ensemble models, CatBoost yielded the most favorable point estimates, with MAE = 0.02933, RMSE = 0.03720, and R² = 0.84958. Strict leave-one-municipality-out validation showed lower zero-shot spatial transferability, with Extra Trees achieving the highest pooled performance (MAE = 0.05198, RMSE = 0.06464, R² = 0.58159). Extra Trees-based local adaptation substantially improved transfer to previously unseen communities, reaching pooled MAE = 0.03234, RMSE = 0.04121, and R² = 0.83424 after 120 days of community-specific observations, although performance remained heterogeneous across municipalities. SHAP analysis identified tourism activity, air temperature, household income, the proportion of the urban population, weekends, and distance to water bodies as important contributors to the model predictions and revealed nonlinear model-specific associations for tourism activity, temperature, and income. Based on the predicted organic waste generation, a scenario-based assessment of the annual bioenergy potential of the investigated communities was performed, demonstrating substantial differences in potential electricity generation among communities. The scientific contribution of this study lies in the integration of ensemble forecasting, explainable artificial intelligence methods, and scenario-based bioenergy potential assessment within a single decision-support framework. The practical value of the proposed approach lies in its potential use for preliminary forecasting of organic waste resource availability and scenario-based assessment of bioenergy potential. Further operational, infrastructure, or investment applications require external validation using independent municipal datasets and verification under real operating conditions. Full article