Search Results (871)

Gender bias in generative artificial intelligence (GenAI) is both a technical and a social phenomenon: it emerges from historically patterned data, model design, and interactions in institutional use, and it cannot be understood by engineering or by social critique alone. This critical integrative review develops a more differentiated account. It connects feminist epistemology, Science and Technology Studies, critical AI scholarship, natural language processing, and governance research to examine five levels: historical knowledge production, technical representation and generation, benchmark evaluation, institutional deployment, and accountability. The review explains tokenization, next-token prediction, transformers, and the transition from static embeddings to contemporary language models before assessing evidence from standard fairness tests—coreference tests (WinoBias), sentence-pair tests (CrowS-Pairs), and stereotype tests (StereoSet)—as well as open-ended generation, multilingual testing, and text-to-image systems. It shows that measured bias varies with task, prompt, language, model version, and metric. What a test records and what that record means are therefore distinct questions: measurements are situated and depend on the instrument, and their interpretation draws on theory rather than following from the numbers alone. Evidence from employment, education, healthcare, and translation further indicates that the relevant unit of analysis is the model-in-context—the model together with the institution and workflow in which its outputs are used. Technical mitigation can reduce specific harms but does not repair unequal criteria, incomplete evidence bases, or weak institutional accountability. The review proposes a multilevel governance approach combining technical evaluation, documentation, professional and community oversight, appeals, remedies, and public-interest knowledge infrastructure. Its distinctive contribution is to connect three observations usually kept apart—how bias is measured, how generative systems concentrate epistemic authority, and how statistical learning is oriented toward past data—and to show why democratic and feminist governance can keep alternative technological futures open. Full article

Flexible pressure sensors featuring high sensitivity, a broad detection range, and excellent stability are pivotal components for high-precision electronic skins and human health monitoring. To circumvent the limitations of existing sensors in maintaining high responsiveness across extensive pressure ranges, herein, a novel flexible capacitive pressure sensor is developed based on a 3D porous ionogel foam composite (IL/EG/PVA@MF) coupled with a planar electrode array. This device leverages the synergistic structural engineering of the 3D hyperelastic melamine foam (MF) skeleton and the pressure-regulated fringe-field distribution and iontronic interfacial polarization of the porous ionogel. Experimental evaluations demonstrate that the sensor achieves a high normalized sensitivity of 62.45 kPa⁻¹ (2–10 kPa) and maintains reliable piecewise linear sensing performance across a broad working range of 0–50 kPa, accompanied by a rapid response time of within 8 ms. Furthermore, the sensor exhibits outstanding performance consistency after 6000 compression-release cycles at 50 kPa, verifying its good mechanical durability. In practical applications, the device can monitor diverse physiological signals with high fidelity, ranging from subtle radial artery pulses to large-scale joint movements and specific coughing patterns, underscoring its broad potential for integrated wearable systems and intelligent healthcare. Full article

Artificial intelligence has emerged as a promising paradigm for advancing the screening, diagnosis support, and monitoring of autism spectrum disorder (ASD) through scalable and data-driven clinical augmentation. Recent advances in machine learning, multimodal sensing, and digital phenotyping have enabled AI systems to analyze behavioral, neurophysiological, speech, and clinical data to identify early markers of ASD. Despite encouraging experimental results, major barriers to clinical translation remain, including limited generalizability, fragmented datasets, insufficient evaluation rigor, lack of semantic interoperability, and unresolved ethical and regulatory concerns. This systematic review provides a comprehensive technical review of AI for ASD, covering data modalities, feature engineering, learning paradigms, evaluation protocols, deployment architectures, and knowledge representation frameworks. Particular emphasis is placed on system-level and translational considerations, including cloud–edge infrastructures, explainable clinical decision-support systems, privacy-aware deployment, and ontology-driven reasoning. Beyond summarizing existing work, this paper critically analyzes challenges related to reproducibility, dataset bias, interpretability, and clinical integration and derives design requirements for next-generation trustworthy ASD AI systems. We argue that meaningful clinical impact will require the integration of multimodal learning, semantic knowledge representation, explainable reasoning, and human-in-the-loop decision processes to support safe, interpretable, and clinically deployable AI systems in pediatric healthcare environments. Full article

The increasing complexity of healthcare data ecosystems demands educational technologies capable of supporting data-intensive learning through advanced analytics, immersive interfaces, and learning analytics. This paper presents a cloud-based reference architecture and a prospective evaluation protocol for integrating Business Intelligence (BI), Extended Reality (XR), and learning analytics in health data science education. The proposed architecture is informed by a systematic literature review conducted according to the PRISMA 2020 guidelines, which screened 613 records retrieved from four databases and retained 56 studies for qualitative synthesis. The review indicates that, although BI and XR technologies have independently been associated with educational benefits, empirical evidence supporting integrated educational architectures combining BI, XR, and learning analytics remains limited, particularly in health data science education. Based on these findings, the paper specifies a layered reference architecture comprising a cloud analytics engine, an immersive visualization engine, an interoperability layer, and a learning analytics pipeline designed to support adaptive and AI-assisted educational services during subsequent implementation phases. The reference architecture is partially instantiated within the curricular unit Health Data Analysis and Visualization of the Digital Health programme at the Polytechnic University of Porto, where the BI and XR components are currently deployed and used within the course, while the interoperability middleware, learning analytics infrastructure, and AI-assisted services remain under development or are specified as architectural capabilities. To support future empirical validation, the paper also defines a comprehensive prospective evaluation protocol comprising predefined outcomes, established instruments with published psychometric properties, together with an expert-developed health data literacy assessment undergoing content validation, research hypotheses, power analysis, a statistical analysis plan, and ethical and data-governance provisions. The manuscript makes four principal research contributions: (i) a cloud-based reference architecture for BI–XR integration, (ii) a computational learning analytics pipeline specification, (iii) an interoperable system design for health data science education, and (iv) a prospective evaluation protocol to guide the future validation of the proposed reference architecture. Full article

Wound management continues to represent a major global healthcare challenge, with the wound care market growing each year and a rising incidence of chronic wounds worldwide. Effective wound healing requires dressings that protect injured tissue, prevent infection, and actively modulate the wound microenvironment to promote tissue regeneration. In recent years, smart polymeric wound dressings have emerged as a functional, more advanced class of wound dressings, engineered from materials capable of responding to stimuli. Physically responsive systems include moisture-adaptive dressings that prevent wound dryness or maceration, pressure-sensitive dressings incorporating flexible capacitive sensors for high mechanical stress mapping, thermoresponsive dressings exploiting sol–gel transitions for temperature-controlled drug release, light-responsive dressings enabling photothermal and photodynamic therapy, and electro-responsive dressings integrating conductive polymers for self-powered electrical stimulation or closed-loop wound monitoring. Chemically responsive systems exploit endogenous biochemical signals, including pH shifts for wound monitoring, reactive oxygen species-cleavable bonds for on-demand drug release, and glucose-responsive platforms for autonomous glycemic regulation in diabetic wounds. Biologically responsive dressings use enzymatic triggers, such as matrix metalloproteinases, hyaluronidase, and bacterial proteases, to achieve autonomous drug delivery. Film-forming sprays further expand the versatility of smart polymeric dressings by enabling contactless application adaptable to irregular wound shapes. In this review, we summarize recent advances in the design, stimuli-responsive mechanisms, characterization methods, and therapeutic outcomes of smart polymeric dressings for wound treatment. Despite promising preclinical results, challenges related to clinical translation, regulatory standardization, and scalable production remain and must be addressed to facilitate widespread clinical adoption. Future directions include multi-stimuli responsive platforms, artificial intelligence-guided wound monitoring, bioprinting of specific dressings, and environmentally sustainable biomaterial design. Full article

Background/Objectives: Health systems involve numerous stakeholders interconnected through nonlinear relationships. While Data Envelopment Analysis (DEA) has been widely used to measure health system efficiency, conventional estimates may exhibit finite-sample bias. An important question, therefore, concerns how health system performance can be measured more reliably, and what factors explain cross-country differences in efficiency. This study introduces an integrated framework that combines Two-Stage Bootstrap DEA with machine learning to assess the performance of the health systems of 26 OECD countries using 2022 data. Methods: In the first step, technical efficiency scores are computed using an output-oriented constant returns to scale (CRS) DEA model. Subsequently, bias-corrected efficiency estimates are derived using the Bootstrap procedure proposed by Simar and Wilson. In the second step, truncated regression analysis and machine learning-based partial dependence analysis, the latter validated through leave-one-out cross-validation, are employed to investigate the determinants of efficiency. Results: The Bootstrap procedure reveals statistically significant differences from conventional DEA results, and bias-corrected results indicate that South Korea, Canada, and the United States achieve the highest efficiency levels. The findings show that tobacco use prevalence has a significantly negative association with health system efficiency and alcohol consumption exhibits a negative, threshold-type pattern, while GDP per capita and out-of-pocket health expenditure display more complex, nonlinear effects. Furthermore, the scenario analysis indicates that a 10% reduction in tobacco use yields the largest predicted single-intervention improvement, while combined interventions produce additional but sub-additive gains. Conclusions: The proposed framework presents a transparent and validated approach for assessing and explaining health system performance, generating findings relevant to the development of evidence-based health policy. Full article

Photocatalytic antibacterial coatings have emerged as a promising antibiotic-free strategy for combating healthcare-associated infections, biofilm formation, marine biofouling, and environmental microbial contamination. Unlike conventional antimicrobial approaches, photocatalytic systems continuously generate reactive oxygen species (ROS) under light irradiation, enabling broad-spectrum antimicrobial activity while minimizing the risk of antimicrobial resistance. This review systematically summarizes the fundamental mechanisms underlying photocatalytic antibacterial activity, including photogenerated charge-carrier dynamics, ROS generation pathways, and microbial inactivation processes. We further highlight recent advances in photocatalyst design, spanning conventional semiconductor photocatalysts, heterojunction engineering, cocatalyst modification, and two-dimensional material-assisted strategies for enhanced photocatalytic performance. Crucially, particular emphasis is placed on coating architectures and interfacial regulation, including encompassing fabrication methodologies, coating–substrate adhesion, internal heterointerface design, and coating–microorganism interactions, which dictate long-term durability and antibacterial efficacy. Finally, we explore the diverse applications of these coatings in medical devices, environmental remediation, and marine antifouling, while identifying current bottlenecks and future research trajectories toward developing durable, highly efficient, and clinically translatable antimicrobial surface technologies. Full article

Central nervous system (CNS) disorders represent a significant healthcare challenge, with aging as the primary risk factor. Current clinical management remains predominantly symptomatic, as late-stage diagnosis and the blood–brain barrier (BBB) limit therapeutic efficacy. This review synthesizes emerging innovations in neurotheranostics—integrated diagnostic and therapeutic platforms—focusing on the neurovascular unit (NVU) as a central pathogenic driver and target. Evidence indicates that NVU and BBB dysfunction are early events in Alzheimer’s, Parkinson’s, amyotrophic lateral sclerosis, and Huntington’s diseases, often preceding classic neuropathological hallmarks. The review highlights the potential of nanotechnology, engineered nanoparticles (NPs) and microRNAs (miRNAs) as precision tools for early detection and targeted CNS delivery. Additionally, it discusses the transformative impact of artificial intelligence (AI) in facilitating personalized, predictive care. Transitioning from a generic “one-pill-for-one-disease” model to a patient-centered strategy targeting early NVU alterations is essential. Integrating AI, nanotechnology and NVU-focused strategies offers a promising path toward effective, personalized disease-modifying therapies. Full article

This paper presents a data-driven educational framework for teaching psychrometry and air quality control in hospitals, developed within an international pilot project involving universities and hospitals in Spain, Romania, and Turkey. The objective of this pilot study is to examine a multidisciplinary framework that equips engineering students with essential technical skills for managing hospital infrastructure, particularly in critical areas like operating rooms and intensive care units. The methodology integrates theoretical instruction, analogue instruments, and digital technologies, including Arduino-based sensing and AI tools, to facilitate data interpretation and critical thinking. By bridging manual measurements with digital monitoring, the framework aims to equalize proficiency among students from diverse engineering backgrounds. Quantitative results from 23 participants provide preliminary evidence of academic growth, consistent with the hypothesis that this integrated approach may facilitate conceptual mastery. This work offers preliminary insights into the advancement of data-driven modelling in engineering education, emphasizing the significance of multidisciplinary training and international collaboration in preparing future professionals for the oversight, operational management, and maintenance of modern healthcare facilities. Full article

Xylitol is a five-carbon sugar alcohol widely used in the food, pharmaceutical, oral healthcare, and personal care industries because of its low caloric value, low glycaemic index, and non-cariogenic properties. Industrial production is mainly based on catalytic hydrogenation of xylose, which provides high conversion efficiency but requires intensive energy input, costly catalysts, and complex purification processes. Microbial fermentation has emerged as a sustainable alternative for producing xylitol from renewable lignocellulosic biomass. This review summarizes recent advances in xylitol production, with a particular focus on microbial biomanufacturing. Key developments in lignocellulosic biomass utilization, metabolic engineering, cofactor balancing, oxygen regulation, and fermentation optimization are discussed. Chemical and biological production routes are critically compared in terms of efficiency, sustainability, and industrial applicability. Recent progress in downstream purification and biorefinery integration is also highlighted. Despite substantial advances, challenges including inhibitor toxicity, limited microbial robustness, low fermentation productivity, and high purification costs continue to hinder large-scale commercialization. Future research should focus on feedstock valorization, systems metabolic engineering, process intensification, and sustainable separation technologies to improve the economic and environmental sustainability of bio-based xylitol production. Full article

Background/Objectives: Healthcare delivery systems are inherently complex, and efforts to improve quality and safety often fail to achieve sustained, system-wide impact because of misalignment between work system design, care processes, and clinical practice. This article presents a practical, theory-informed framework that integrates human factors and systems thinking to guide the design and implementation of clinical pathways for reliable care delivery and system-level transformation. Methods: A conceptual framework was developed through synthesis of literature from human factors engineering, sociotechnical systems theory, and healthcare quality improvement. The framework organizes healthcare delivery into three interconnected domains—the work system, care processes, and outcomes—while positioning clinical pathways as the operational mechanism linking system design to care execution. Pediatric immunization was used as an illustrative case example. Results: The framework illustrates how misalignment in system design may contribute to variability in care processes and outcomes. Clinical pathways integrated into routine workflows and supported by human factors strategies—including decision-support heuristics, workflow simulation, and structured debriefing—may promote more reliable care delivery. Sustained implementation across teams and clinical settings is expected to reduce missed opportunities, improve consistency of care, and support progression toward system-level transformation. Conclusions: Clinical pathways can serve as practical tools for translating system design principles into reliable healthcare delivery when implemented using human factors and systems-based approaches. This framework provides healthcare leaders and quality improvement teams with a structured approach for designing, implementing, and scaling pathways to support safer, more reliable, and more consistent care delivery. Full article

The Central Sterile Supply Department (CSSD) functions as the critical control point for hospital infection prevention, yet its equipment-intensive environment generates occupational noise exposure that has been systematically overlooked in both the healthcare quality and occupational health literature. We conducted a focused narrative review across PubMed, Web of Science, and Scopus (up to March 2026) to examine the prevalence, psychophysiological mechanisms, and management of noise in CSSD settings. Four peer-reviewed studies investigating CSSD noise exposure were identified, conducted in Chinese and Brazilian hospitals. The evidence indicates that air guns and pressure steam sterilizers constitute the dominant noise sources, with self-reported exposure associated with psychological symptoms and sleep disturbance in approximately one-quarter of staff. Notably, health concerns may mediate a substantial proportion of the noise–psychology relationship, suggesting a cognitive–affective pathway that may be as theoretically significant as direct physiological damage, although this cross-sectional finding requires longitudinal confirmation. No study has empirically linked CSSD noise to sterilization failures, device damage, or patient-level outcomes. Drawing on this nascent evidence base and adjacent occupational noise theory, we advance three testable propositions: that CSSD noise may constitute a systemic quality risk factor, that subjective cognitive–affective appraisal plays a primary mediating role, and that effective management requires multi-level integration across engineering, administrative, and psychological domains. This review reframes CSSD noise from an ergonomic nuisance to an embedded environmental stressor and charts a structured research agenda for objective measurement, mechanism validation, and intervention evaluation. Full article

Background/Objectives: ICU nurse staffing Standard Operating Procedures (SOPs) govern nurse-to-patient assignment and escalation rules, yet their robustness under demand variability and workforce disruption has not been quantitatively evaluated. Methods: A Discrete Event Simulation model of a 20-bed ICU was parameterized from 20,419 MICU stays extracted from MIMIC-IV v3.1. Three SOPs—Fixed Ratio, Acuity-Adjusted, and Dynamic Escalation—were evaluated across four scenarios (baseline, census surge, nurse shortage, combined disruption), each with 100 replications, using Kruskal–Wallis and Dunn’s post hoc tests. Results: Dynamic Escalation achieved the lowest adverse event rate at baseline (4.25 per 100 patient days), a 16.8% reduction relative to Fixed Ratio (5.11). However, a resource-constrained comparison equalizing total nurse hours across protocols reversed this ranking: Dynamic Escalation performed significantly worse than both static protocols (p < 0.0001), which were themselves statistically indistinguishable, indicating that its primary advantage stems from greater staffing capacity rather than superior decision logic. Conclusions: Apparent differences between ICU staffing SOPs are largely driven by total available nurse hours rather than allocation logic. Investment in staffing capacity, rather than a specific allocation rule, may be the primary lever for improving ICU patient safety outcomes. Full article

Modern societies comprise overlapping communities whose opinions evolve on strongly interacting networks that are often mutually antagonistic. We introduce a minimal antagonistic multiplex consensus model in which each layer follows intra-layer majority-rule dynamics, while inter-layer interactions are inhibitory. A mean-field analysis shows that antagonistic coupling destabilizes the balanced state through an antisymmetric mode and favors two polarized absorbing states with opposite magnetization in the two layers. Network-averaged simulations confirm that small fluctuations near equal initial support determine which polarized state is ultimately reached: trajectories exhibit metastable delay, long convergence times, and a localized peak in the Shannon entropy of outcomes. A finite-size analysis with independent network realizations and bootstrap uncertainty estimates shows that the high-entropy interval narrows as $\Delta r$ $\sim N^{-{\gamma }_{\mathrm{eff}}}$, with ${\gamma }_{\mathrm{eff}}=0.513$ and a 95% bootstrap confidence interval $[0.489,0.526]$, consistent with finite-size sharpening controlled by fluctuations in the initial imbalance. We also perform network topology checks and find that the qualitatively antagonistic mechanism persists beyond random-regular graphs. As an illustrative empirical application, we analyze county-level results from the 2024 U.S. presidential election. The vote-share and entropy landscapes separate low-entropy partisan strongholds from higher-entropy competitive counties. Our results suggest that antagonistic multiplex coupling provides a simple mechanism by which polarized attractors and localized outcome uncertainty can arise together. Full article