Search Results (1,451)

In multi-turn conversational AI, individually innocuous personally identifiable information (PII) fragments disclosed across successive turns can accumulate into a re-identification risk that no single utterance reveals on its own. Existing PII detectors operate on isolated utterances and therefore cannot track this cross-turn evidence build-up. We propose a stateful middleware guardrail whose core design principle is speaker-attributed entity isolation: every extracted PII fragment is attributed to its originating conversational participant, and evidence is accumulated in entity-isolated subgraphs that prevent cross-entity contamination. The system signals re-identification onset $t_{\mathrm{pred}}$ at the earliest turn where combination-based rules grounded in the uniqueness literature are satisfied. On a 184-record template-synthetic evaluation corpus, the gated NER configuration leads on primary timeliness (OW@5 $= 73.4\%$, MAE$= 1.357$ turns); the full system achieves OW@5 $= 70.7\%$ with MAE $= 2.442$ turns as an alternative operating mode for ambiguity-sensitive disclosure patterns. We further evaluate behavior on a 300-record mutation stress set, test RULE_B on the ABCD external corpus, and supplement RULE_A evaluation with both a proxy-labeled transfer analysis on PersonaChat and a manual annotation study on 151 Switchboard dialogues. The reported results should be interpreted as an initial empirical reference point rather than a sufficient endpoint for autonomous runtime enforcement. Full article

Background: Ankle foot orthosis–footwear combination (AFO–FC) tuning involves structured adjustment of the AFO relative to footwear to optimise shank alignment and ground reaction force (GRF) positioning during stance. Although established biomechanical frameworks and clinical algorithms are available, variability in clinical implementation persists. Previous investigations have primarily relied on self-reported practice within single healthcare settings and have not, to our knowledge, systematically examined how orthotists articulate and apply tuning principles within structured clinical reasoning across diverse educational and practice environments. Objectives: This study aimed to determine the level of awareness and framework-based proficiency in AFO–FC tuning among practising orthotists in a geographically diverse convenience sample, to examine the extent to which AFO–FC tuning is integrated into routine clinical practice, and to explore associations between framework-based proficiency level and selected professional characteristics. Methods: A cross-sectional study was conducted using an online survey of practising orthotists (n = 245). Awareness of AFO–FC tuning and self-reported routine implementation were assessed. Framework-based proficiency was evaluated among respondents reporting awareness (n = 212) using structured content analysis of open-text responses within a predefined exploratory five-domain biomechanical framework, and classified as limited (0–1 domains), partial (2–3 domains), or full (4–5 domains). Associations between framework-based proficiency level and professional characteristics were examined using chi-square tests. Binary logistic regression was performed to assess the association between framework-based proficiency level and self-reported routine implementation. Results: Self-reported awareness of AFO–FC tuning was high (86.5%), whereas 53.5% reported routine implementation. Based on the framework scoring, 59.0% demonstrated limited framework-based proficiency, 31.6% partial framework-based proficiency, and 9.4% full framework-based proficiency. No statistically significant associations were observed in this sample between framework-based proficiency level and educational qualification, years of clinical experience, or annual AFO case volume (p > 0.05). Full framework-based proficiency was associated with higher odds of self-reported routine implementation (OR = 4.03, 95% CI 1.44–11.25, p = 0.008). Conclusions: Despite high self-reported awareness, framework-based proficiency in AFO–FC tuning was limited within this sample. Self-reported routine implementation was more frequently reported among respondents with higher framework-based proficiency, whereas no statistically significant associations were observed with educational level, clinical experience, or annual AFO case volume. These hypothesis-generating findings should be interpreted cautiously given the cross-sectional design and framework-based (non-validated) classification. Full article

Artificial intelligence (AI) is increasingly used to support tuberculosis (TB) screening and diagnosis, particularly through computer-aided detection (CAD) applied to chest radiography (CXR). However, the programmatic value of AI depends not only on diagnostic accuracy but also on implementation context, threshold calibration, and integration into diagnostic pathways. We conducted a narrative, state-of-the-art review of AI applications across the TB diagnosis pathway. Evidence was synthesized from World Health Organization policy documents, independent validation initiatives, and peer-reviewed studies published between 2010 and 2026, with a structured selection process aligned with PRISMA principles. CAD for CXR is the most mature AI application and is recommended by WHO for TB screening and triage among individuals aged ≥15 years in specific contexts. Across studies, CAD-CXR demonstrates sensitivity comparable to human readers, although performance varies by product, population, and imaging conditions, necessitating local threshold calibration. Evidence from implementation studies suggests improvements in screening efficiency and potential cost-effectiveness in high-burden settings. Other AI modalities, including computed tomography (CT)-based imaging analysis, point-of-care ultrasound interpretation, cough or stethoscope sound analysis, clinical risk models, and genomic resistance prediction show promising but heterogeneous results, with most requiring further independent validation and prospective evaluation. AI has the potential to strengthen TB screening and diagnostic pathways, but its impact depends on integration into health systems and evaluated using patient- and program-level outcomes rather than accuracy alone. A differentiated approach is needed, with responsible scale-up of policy-endorsed tools alongside rigorous evaluation of emerging technologies to support effective and equitable TB care. Full article

Cultured meat holds the potential to reduce environmental impacts and offer ethical advantages while replicating the nutritional, taste, and texture attributes of conventional meat. To date, most research on consumer acceptance of meat has focused on European and North American markets. In contrast, Muslim-majority countries remain underexplored, particularly regarding the compatibility of cultured meat with Islamic dietary laws. These societies are experiencing rising meat consumption, and countries such as Saudi Arabia and Malaysia rely heavily on meat imports. This narrative perspective article aims to systematically examine how specific stages of cultured meat production align with, or challenge, Islamic dietary (halal) principles. To this end, we adopt a stage-based analytical approach, mapping key technological steps in cultured meat production onto core requirements of Islamic jurisprudence. To this end, we critically and comprehensively examine the intersection between cultured meat production methods and the Islamic concept of halal, which extends beyond ingredient permissibility to encompass ethical, spiritual, and hygienic dimensions of food production. Key challenges to halal certification include the origin and status of starter cells, whether donor animals were slaughtered according to Islamic law, the permissibility of biopsied tissue, and the use of fetal bovine serum in growth media. The analysis indicates that while halal-compliant cultured meat is scientifically feasible, its adoption remains constrained by unresolved religious interpretations, regulatory fragmentation, and limited availability of halal-certified inputs. We emphasize the need for interdisciplinary collaboration among Islamic scholars, food scientists, certification bodies, and policymakers. From a policy perspective, harmonized halal standards, targeted investment in serum-free and animal-free culture media, and early regulatory engagement with Islamic authorities are essential to facilitate responsible market entry. Therefore, we suggest a multi-level governance and stage-gated halal decision framework for cultured meat. Proactive regulation and open dialogue with religious leaders are vital to ethically introduce cultured meat into Muslim markets, aligning innovation with Islamic values while supporting national sustainability and food security goals. Full article

This study examines the design and pedagogical evaluation of Extended Reality (XR) applications, with a primary focus on location-based Augmented Reality (AR). The XR applications were implemented within an environmental education program delivered by the Education Center for the Environment and Sustainability (E.S.E.C.) of Kastoria, aiming to enhance students’ understanding of lake ecosystems and environmental awareness through immersive, situated learning experiences. The development followed the ADDIE instructional design framework and was grounded in principles of experiential and situated learning. The educational intervention was conducted in an authentic field setting along the shoreline of Lake Kastoria and combined location-based AR activities with complementary immersive VR experiences. Evaluation data were collected through a structured questionnaire administered to 271 primary and secondary school students, employing XR-relevant constructs including Challenge/Satisfaction/Enjoyment, Ease of Use, Usefulness/Knowledge, Experiential and Situated Learning, Interaction/Collaboration, and Intention to Reuse. In addition, accompanying teachers provided supplementary qualitative feedback to support the interpretation of the findings under authentic field conditions. Descriptive statistical analysis indicated consistently high scores across all constructs (M = 3.27–4.40, SD = 0.41–0.64). Pearson correlation analysis revealed strong associations between Experiential/Situated Learning and Usefulness/Knowledge (r = 0.737), Experiential/Situated Learning and Challenge/Satisfaction/Enjoyment (r = 0.642), Intention to Reuse and Challenge/Satisfaction/Enjoyment (r = 0.635), as well as Usefulness/Knowledge and Challenge/Satisfaction/Enjoyment (r = 0.619). Multiple regression analyses further supported key relationships, including Usefulness/Knowledge as a predictor of Experiential/Situated Learning (β = 0.57, p < 0.001), Experiential/Situated Learning as a predictor of Challenge/Satisfaction/Enjoyment (β = 0.47, p < 0.001), and Interaction/Collaboration as a predictor of Intention to Reuse (β = 0.31, p < 0.001). Intention to reuse was mainly associated with interaction and collaboration, enjoyment and motivation, perceived usefulness/knowledge, and ease of use. Overall, the findings indicate that XR-supported outdoor learning is positively associated with key experiential, emotional, social, and perceived learning dimensions when embedded within a coherent pedagogical framework. Full article

In the field of engineering, the advancement of automated scoring systems for laboratory reports has been significantly hampered by three persistent challenges: scarcity of high-quality annotated data, high domain-specific complexity, and insufficient model interpretability. To address these limitations, this study proposes an AdaBoost regression model based on multi-level feature engineering and threshold control, denoted as MFTC-ABR. This method constructs a multi-dimensional feature set using a lightweight neural network, which evaluates laboratory reports across four core dimensions: comprehension of experimental principles, completion of experimental procedures, depth of result analysis, and plagiarism detection. At the scoring algorithm level, a dynamic threshold adjustment mechanism is integrated into the AdaBoostReg ensemble learning framework. By redesigning the sample weight update rule, the prediction errors of samples are divided into three intervals: the acceptable region, the stable learning range, and the focus range. Accordingly, a differentiated weight update strategy is implemented, and a history-aware mechanism is introduced to further regulate the attention allocated to individual samples. Finally, experimental results on the power electronics laboratory report dataset show that MFTC-ABR model achieves a mean absolute error (MAE) of 3.09 and a scoring consistency rate of 82% within a five-point error tolerance. These findings validate the effectiveness and practicability of the proposed method for automatic assessment in specialized domains with limited data availability. Full article

Feature weighting plays a central role in medical classification by enhancing predictive accuracy, interpretability, and clinical trust through the explicit quantification of variable relevance. Despite their widespread use, existing filter-, wrapper-, and embedded-based feature weighting methods are predominantly deterministic and exhibit pronounced sensitivity to label noise and outliers, which are pervasive in real-world medical data. This often results in unstable importance estimates and unreliable clinical interpretations. In this work, we introduce a novel Bayesian feature weighting model that fundamentally departs from existing approaches by jointly integrating simplex-constrained Dirichlet priors for global feature weights, hierarchical shrinkage priors for coefficient regularization, and contamination-aware priors for explicit modeling of label noise within a single coherent probabilistic framework. Unlike conventional Bayesian feature selection or robust classification models, the proposed formulation yields globally interpretable feature weights defined on the probability simplex, while simultaneously providing full posterior uncertainty quantification and robustness to both mislabeled observations and aberrant feature values through principled influence control. Comprehensive simulation studies across diverse contamination scenarios, together with applications to multiple real-world medical datasets, demonstrate that the proposed model consistently outperforms classical and state-of-the-art baselines in terms of discrimination, probabilistic calibration, and stability of feature-importance estimates. These results highlight the practical and methodological significance of the proposed framework as a robust, uncertainty-aware, and interpretable solution for medical decision making under noisy data conditions. Full article

Generative artificial intelligence in higher education is typically framed as either a student productivity tool or an institutional disruption. This agenda-setting essay advances a third position: mobile generative AI functions as relational infrastructure—a persistent communicative presence that mediates identity, meaning-making, and belonging during institutional transition. Focusing on international graduate student onboarding, we abductively “think through” two complementary theoretical lenses. Constitutive Artificial Intelligence Identity Theory (CAIIT) conceptualizes AI as a co-constitutive participant in identity formation through recursive communicative feedback loops. Language Convergence/Meaning Divergence (LC/MD) theory explains how shared institutional language masks interpretive gaps across intercultural and bureaucratic contexts. Reading narrative vignettes through these frameworks, we argue that generative AI is neither simple curricular tool nor personal aid, but both relational and organizational infrastructure, redistributing translational, emotional, and interpretive labor in higher education. We outline four design principles for AI-integrated onboarding: distinguish communicative scaffolding from cognitive replacement; design systems that assume meaning divergence; center equity in AI-mediated transitions; and anticipate ethical risk. Reframing AI as relational infrastructure shifts AI-in-education research toward relational accountability and institutional care. Full article

Background: Severe obesity is associated with chronic low-grade inflammation, dysglycemia, and higher periodontitis risk. Sleeve gastrectomy (SG) is now a dominant bariatric procedure and reliably improves weight and metabolic status yet reported oral and periodontal trajectories after surgery remain heterogeneous. Objective: To synthesize SG-centered evidence on periodontal outcomes, oral and gut microbiome remodeling, and mechanistic pathways that may link postoperative physiology to the gut–oral axis. Methods: We conducted a structured narrative review guided by SANRA principles using targeted searches of PubMed/MEDLINE, Web of Science, Scopus, and Embase, complemented by citation chaining of key reviews and mechanistic anchor papers; evidence was organized into clinical, oral microbiome, gut microbiome, and mechanistic gut–oral axis streams and interpreted with a pragmatic evidence hierarchy. Results: Small prospective SG cohorts suggest bleeding on probing (BOP), gingival indices, and sometimes probing depth (PD) may improve in some patients, particularly alongside weight loss, improved glycemic control, and lower systemic inflammatory burden, whereas clinical attachment level (CAL) and longer-term structural trajectories remain mixed; mixed-procedure syntheses also report early deterioration in some settings. Oral microbiome findings after bariatric surgery are site- and time-dependent, and salivary signals do not necessarily mirror subgingival plaque, whereas gut microbiome remodeling and bile acid signaling changes are more consistently reported and provide plausible but indirect mediator candidates. At the same time, reflux, vomiting, salivary changes, diet patterning, medications, and periodontal care can modify or counteract potential periodontal benefits and may increase competing risks such as caries or erosive tooth wear. Conclusions: The SG–gut–oral axis-periodontal pathway is a biologically plausible working hypothesis rather than a proven causal pathway in humans. The present evidence for any periodontal benefit relies mainly on small observational cohorts and is most credibly demonstrated for inflammatory, not structural, endpoints. Full article

NetworkGuard is a modular edge-based virtual network sensing framework designed for residential smart home security. The system interprets network telemetry—such as DNS queries, firewall events, VPN latency, and connection establishment delay—as structured sensing signals for gateway-level monitoring. Implemented on a Raspberry Pi 4 and managed via an Android interface, NetworkGuard integrates DNS filtering (Pi-hole), firewall enforcement (UFW), encrypted VPN tunneling (WireGuard), and an AI-assisted advisory layer for contextual log interpretation. During a six-week residential deployment, DNS blocking efficiency improved from 81.2% to 97.0% following blocklist refinement, while VPN connection establishment time decreased from approximately 3012 ms to 2410 ms after configuration tuning. ICMP-based measurements indicated a stable tunnel latency under moderate traffic conditions. Controlled validation scenarios—including DNS manipulation attempts, port scanning, and VPN interruption testing—confirmed consistent firewall enforcement and tunnel containment. The results demonstrate that layered security principles can be adapted into a lightweight, reproducible edge architecture suitable for small-scale residential IoT environments without a reliance on enterprise infrastructure. Full article

Prior to the Tang Dynasty, interpretations of the Laozi fell into two traditions: the Pre-Qin and Han tradition, represented by Yan Zun, and the Wei–Jin tradition, represented by Wang Bi. The commentaries on the Laozi in circulation today are influenced by metaphysics in emphasizing “non-being” (wu) as the substance of the Dao (dao). Yan Zun’s Laozi zhigui 老子指歸 (lit. “Purport of the Laozi”) is the oldest extant commentary. In his thought, Yan carries on the legacies of the Laozi and the Zhuangzi and serves as a precursor to later religious Daoism. Yan Zun established a triadic framework—comprising the Dao, Vacuity, and Spontaneity—that shaped Han and Tang Daoism. This reading inherits the Pre-Qin Daoist principle that takes Vacuity as its ontological root and yielding softness as its operative function, laying the theoretical foundation for religious Daoist thought in the Jin and Tang dynasties. Yan Zun’s interpretations of the Laozi frequently surprise modern scholars, yet his views align closely with the contents of the Mawangdui Laozi silk manuscripts (c. 168 BCE) and Peking University Western Han bamboo-slip Laozi (c. 150 BCE), which demonstrates his distinctive scholarly contribution and contemporary relevance. Full article

Detecting unnatural pauses is a critical component of automated quality assessment (AQA) in interpreter training, as pause patterns directly reflect an interpreter’s cognitive load and fluency. Traditional pause detection methods rely on static temporal thresholds (e.g., 1.0 s), which often fail to account for segment-specific speech rate variability and individual speaking styles. This study proposes a context-adaptive pause detection framework that integrates unsupervised anomaly detection using Isolation Forest (iForest) with a sliding window technique. To enhance pedagogical validity, we specifically focused on intra-sentential pauses by delineating sentence boundaries using a specialized segmentation model. The proposed model was evaluated against ground-truth labels annotated by professional interpreting experts. Our results demonstrate that the sliding window–based contextual anomaly detection model significantly outperforms the conventional static baseline, particularly in terms of recall and Cohen’s kappa. Furthermore, by applying a weighted F3-score and the “Recognition-over-Recall” principle, we confirmed that the proposed model substantially reduces the instructor’s total operational burden by shifting the workload from de novo annotation creation to more efficient corrective pruning. These findings suggest that speech-adaptive modeling provides a more reliable and labor-saving framework for automated interpreting assessment and feedback. Specifically, this study makes three main contributions: (1) the proposal of a context-adaptive pause detection framework using anomaly detection, (2) the integration of sliding window–based local contextual modeling for speech-rate–aware analysis, and (3) the introduction of an evaluation strategy based on the Recognition-over-Recall principle to reduce instructor workload in interpreter training. Full article

Background/Objectives: Although prescribing physical activity (PA) is a well-established preventive strategy in primary health care (PHC), its ethical implications remain under-researched. This study examines how general practitioners (GPs) and nurses experience, interpret, and manage ethical tensions in PAP. Methods: A qualitative study was conducted with 28 PHC professionals (13 GPs, 15 nurses) from rural and urban centers in Toledo, Spain (M = 18.4 years of experience). Data were collected through semi-structured interviews and analyzed using reflexive thematic analysis. Beauchamp and Childress’ four-principles framework was applied abductively to synthesize ethical conflicts and coping strategies. Results: Two main themes emerged: (1) Ethical conflicts in PAP, characterized by tensions between autonomy and paternalism, and the challenge of balancing beneficence with non-maleficence under institutional pressures; and (2) Professional coping strategies, where clinicians used relational care, individualized tailoring, and interprofessional collaboration to mitigate moral distress. Results indicated that clinical codes, such as “unrealistic goals” or “institutional pressure,” often overlapped across multiple ethical principles, necessitating a nuanced, multi-dimensional approach to counseling. Conclusions: PAP is not a neutral clinical task but an ethically grounded practice constrained by structural and organizational factors. To move toward safe and equitable health promotion, PAP must be conceptualized as a relational intervention. We propose an Ethical Reflective Tool and a conceptual framework to support clinical reflection, enhance professional accountability, and guide policy-level support for preventive care in PHC. Full article

Financial fraud and money laundering continue to challenge financial stability and regulatory oversight, motivating the widespread adoption of machine learning models for transaction monitoring. Although ensemble models such as Random Forest and XGBoost achieve strong predictive performance, their deployment in high-stakes financial environments is constrained by limited interpretability, overconfident predictions, and the absence of principled mechanisms for expressing decision uncertainty. Emerging regulatory expectations increasingly emphasise transparency, accountability, and operational reliability, underscoring the need for evaluation frameworks that extend beyond predictive accuracy. This study proposes the Integrated Transparency and Confidence Framework (ITCF), a deployment-oriented approach that unifies model explainability, statistically valid uncertainty quantification, and operational decision support for fraud detection. ITCF combines instance-level explanations generated via Local Interpretable Model-Agnostic Explanations (LIME) with distribution-free uncertainty estimation using split conformal prediction. The framework incorporates selective explainability, abstention-based routing, and uncertainty-driven triage to support human-in-the-loop review. Using the PaySim dataset of 6,362,620 mobile-money transactions, Random Forest and XGBoost models are evaluated under extreme class imbalance using F1-score, AUC–ROC, and Matthews Correlation Coefficient (MCC). At a target coverage level of 90% ($\alpha =0.1$), both models achieve empirical coverage close to the target level, with XGBoost producing smaller prediction sets and superior recall, MCC, and latency. ITCF provides transaction-level explanations for uncertain cases and specifies an auditable workflow that is intended to support transparency, traceability, and risk-aware human review, thereby enabling defensible human decision-making in regulated environments. Overall, this study illustrates how explainability and uncertainty quantification can be combined in a deployment-oriented evaluation workflow while noting that real-world validation remains a future endeavour. Full article

Background: Colorectal cancer (CRC) remains the third most prevalent malignancy globally, with approximately 154,000 new cases and 54,000 projected deaths anticipated for 2025. The recent advancement of foundation models in computational pathology has been largely propelled by task-agnostic methodologies that overlook organ-specific crucial morphological patterns that represent distinct biological processes that fundamentally influence tumor behavior, therapeutic response, and outcomes. Methods: In this study, we develop a novel, interpretable AI model, PRISM (Prognostic Representation of Integrated Spatial Morphology), that incorporates a continuous variability spectrum within each distinct morphology to reflect the principle that malignant transformation occurs through incremental evolutionary processes. PRISM is trained on 15 million histological images extracted from surgical resection specimens of 2957 patients. Results: PRISM achieved superior prognostic performance for five-year OS (AUC = 0.70 ± 0.04; accuracy = 68.37% ± 4.75%; HR = 3.21, 95% CI = 2.18–4.72; p < 0.0001 using multi-variate cox-proportional hazards model), outperforming existing CRC-specific methods by 15% and AI foundation models by ~23% accuracy. It showed sex-agnostic robustness (AUC Δ = 0.02; accuracy Δ = 0.15%) and stable performance across clinicopathological subgroups, with minimal accuracy fluctuation (Δ = 1.44%) between 5FU/LV and CPT-11/5FU/LV regimens, replicating the Alliance cohort finding of no survival difference between treatments. Conclusions: These results establish PRISM as a promising, interpretable tool for AI-driven prognostication, with potential for future extension to other cancer types and stages Full article