Search Results (13,773)

Background/Objectives: Infraocclusion in pediatric patients may result in space loss, malocclusion and the need for complex orthodontic treatment if not detected early. Conventional diagnosis may be subject to human error and can be challenging, particularly in pediatric cases. The aim of this study is to design and evaluate a lightweight, two-stage deep learning framework with integrated explainable AI (XAI) techniques for automated infraocclusion detection in pediatric panoramic radiographs. Methods: Annotated panoramic radiographs of pediatric patients aged 7–11 years were used for training and validation. In the first stage, a MobileNet V2 Lite model was used to detect the region of interest (ROI) comprising premolars and molars. In the second stage, a custom CNN classifier was proposed to distinguish between infraocclusion and no infraocclusion. Model performance was evaluated in terms of diagnostic accuracy, computational complexity, and statistical significance. XAI techniques were also incorporated to visualize model attention and enhance interpretability. Results: The detection stage achieved high reliability with a precision, recall, F1-score, and AP50 values of 0.99, and an AP75 of 0.89, indicating accurate ROI localization. The classification stage reached an overall accuracy of 98.78%, with class-specific accuracies of 99.25% for infraocclusion and 98.31% for no infraocclusion cases. The framework also demonstrated computational efficiency, requiring only 1.88 M trainable parameters (7.19 MB), with short training times and low inference latency (0.8 ms for classification and 19 ms for detection). XAI visualizations consistently highlighted clinically relevant regions, such as occlusal margins and interproximal areas, confirming the model’s alignment with radiographic features recognized by clinicians. Conclusions: The proposed two-stage framework provides an accurate, computationally efficient, and interpretable solution for automated infraocclusion detection in pediatric patients. Its modular design and reduced complexity support practical integration into routine clinical workflows, including resource-constrained environments. These findings indicate that lightweight and XAI systems may enhance early infraocclusion detection while maintaining clinical transparency. Full article

The integration of artificial intelligence (AI) in literature reviews aims to transform research by potentially automating processes, enhancing rigour, and improving quality. The study proposes a structured step-by-step approach to integrate AI tools into the literature review synthesis process. The developed methodological approach has five steps. The first step, planning and readiness, involves scoping, understanding practices, and defining boundaries of AI use. Next is selecting AI tools and aligning their capabilities with the literature needs through a matrix. The third step focuses on using AI to conduct the review, followed by validation and cross-referencing of AI-generated results. The final step is disclosing AI use in line with ethical and reporting standards. The approach is demonstrated through five scenarios: emerging or fragmented literature, large or saturated fields, interdisciplinary domains, methodologically diverse studies, and under-researched topics. This approach is designed to enhance transparency, potentially reduce bias, and support reproducibility by aligning AI functions with research goals. It also addresses ethical considerations and promotes human–AI collaboration. For researchers and academics, it aims to provide a practical roadmap for the responsible adoption of AI in literature reviews, supporting efficiency, ethical tool use, transparency, and the balance between machine assistance and academic judgment. Full article

Background: Artificial intelligence is emerging as a promising tool in surgical oncology, with growing evidence suggesting potential applications in diagnostic support, intraoperative guidance, and perioperative risk assessment. In gastric cancer surgery, emerging applications range from AI-assisted endoscopic detection to data-driven perioperative risk prediction, while some technological developments, particularly in robotic autonomy, derive from broader surgical or experimental models that may inform future gastric procedures. Methods: A narrative review was conducted following established methodological standards, including the Scale for the Assessment of Narrative Review Articles (SANRA) and the Search–Appraisal–Synthesis–Analysis (SALSA) framework. English-language studies indexed in PubMed, Scopus, Embase, and Web of Science up to October 2025 were included. Evidence was synthesized thematically across five domains: AI-assisted anatomical recognition and lymphadenectomy support, autonomous robotic systems, early cancer detection, perioperative predictive and frailty models, and ethical and regulatory considerations. Results: AI-based computer vision and deep learning algorithms have demonstrated promising capabilities for real-time anatomical recognition, surgical phase classification, and intraoperative guidance, although evidence of direct patient-level benefit remains limited. In diagnostic settings, AI-assisted endoscopy and Raman spectroscopy have been shown to improve early lesion detection and reduce dependence on operator experience. Predictive models, including MySurgeryRisk and AI-driven frailty assessments, may support individualized prehabilitation planning and perioperative risk stratification. Persistent limitations include small and heterogeneous datasets, insufficient external validation, and unresolved concerns related to data privacy, algorithmic interpretability, and medico-legal responsibility. Conclusions: Artificial intelligence is progressively emerging as a promising tool in gastric cancer surgery, integrating automation, advanced analytics, and human clinical reasoning. Its safe and ethical adoption requires robust validation, transparent governance, and continuous surgeon oversight. When developed within human-centered and ethically grounded frameworks, AI can augment, rather than replace, surgical expertise, potentially advancing precision, safety, and equity in oncologic care. Full article

Gastronomic tourism heritage represents a significant segment of intangible cultural heritage, reflecting traditional knowledge, local identity, and long-standing culinary practices. The contemporary development of digital technologies, particularly artificial intelligence (AI), opens new possibilities for its preservation, documentation, and sustainable interpretation within cultural tourism. The aim of this research is to examine the role of artificial intelligence as a tool for preserving gastronomic tourism heritage from the perspective of local community members in Bosnia and Herzegovina, Serbia, and North Macedonia, regions characterised by shared gastronomic and cultural traditions. The study was conducted using a quantitative research design based on a structured questionnaire administered to 571 respondents. A convenience sampling approach was applied, targeting individuals involved in the preparation, transmission, or promotion of traditional gastronomy. Data were collected through a combination of field-based and online survey distribution. The analysis focuses on respondents’ perceptions of AI applications in documenting traditional recipes, interpreting gastronomic heritage, and promoting it within tourism, as well as on attitudes related to authenticity and cultural identity preservation. The findings indicate that, within the surveyed sample, artificial intelligence is generally perceived as a useful tool for safeguarding gastronomic heritage. At the same time, respondents emphasise the importance of transparent governance, community participation, and culturally sensitive implementation in order to minimise risks of commodification and loss of authenticity. Full article

Building code waiver assessments in Singapore remain largely discretionary, relying on case officers’ subjective judgement with limited decision-support tooling. This study presents the first machine learning framework for predicting building code waiver outcomes, trained on 197 historically decided cases from the Building and Construction Authority (BCA) across five waiver categories: barrier-free accessibility (n = 45), ventilation (n = 61), staircase design (n = 37), safety provisions (n = 30), and structural modifications (n = 24), spanning 2021 to 2023. Fourteen engineered features, including documentation completeness, technical justification quality, and compliance history, were extracted through domain-expert annotation. Four models were evaluated: L2-regularised logistic regression, random forest, gradient boosting (XGBoost 2.0.1), and a weighted ensemble. The ensemble achieved the highest predictive accuracy of 83.7% (95% CI: 79.2–88.1%) with an area under the receiver operating characteristic curve (AUC) of 0.891 (95% CI: 0.854–0.928), significantly outperforming all individual models (McNemar’s test, p < 0.05). SHAP analysis revealed that documentation completeness and technical justification quality collectively account for 55% of prediction variance. A companion five-by-five risk assessment matrix, combining predicted rejection probability with consequence severity, stratified cases into actionable risk tiers correlating with observed approval rates ranging from 90.3% (very low risk) to 10.0% (very high risk; Spearman rho = −0.71, p < 0.001). Performance varied across waiver categories: ventilation waivers achieved the highest balanced accuracy (87.1%) while safety waivers proved most challenging (balanced accuracy 64.3%, sensitivity 40.0%). The framework offers a transparent, data-driven decision-support complement to regulatory judgement, learning patterns from historically decided applications within the 2021–2023 BCA context, and demonstrates feasibility for integration into Singapore’s Corenet X digital building submission platform. These five waiver categories serve as domain stratification variables. The machine learning target variable is the binary regulatory outcome: Approved (46.2% of cases) or Rejected (53.8%). Full article

Given the global aging of the population and the rising retirement age, the development of cross-generational technologies is crucial for a sustainable workforce supply. While AI-powered smart glasses can provide continuous cognitive support, current industrial solutions often prioritize work efficiency at the expense of the physical, cognitive, and socio-emotional needs of older workers. This study employed a mixed-methods approach grounded in embodied cognition. First, semi-structured interviews with ten participants were analyzed using grounded theory to develop a four-dimensional model of embodied experience: Perceived Pressure, Action Feedback, Collaboration Embedding, and Belonging. Subsequently, four interaction strategies—Rhythm Control, Transparent Feedback, Non-intrusive Assistance, and Legible Privacy & Social Signaling—were formulated and implemented. A high-fidelity prototype was developed to embody these strategies. Finally, a team of eight multidisciplinary experts evaluated the device using the System Usability Scale (SUS) and a proprietary twelve-item questionnaire. The results showed that the device’s overall usability was borderline acceptable (SUS = 68.13 ± 8.94). While the devices received stronger ratings for Control & Safety, the ratings for dignity and social acceptance were comparatively low. These findings contribute to the development of wearable device operation strategies suitable for users of different generations, and underline the importance of social and emotional compatibility as a prerequisite for future practice tests. Full article

In Canada, Indigenous communities have been disproportionately flooded. Specifically, Fort Albany First Nation (FN) located on a flood plain near the mouth of the Albany River in subarctic Ontario, Canada, has been evacuated frequently due to flooding or the threat of flooding―even though dikes were constructed in the late 1990s to safeguard the community. Thus, a fundamental question needs to be asked: Why is Fort Albany FN located on a flood plain in the first place? We answer the question through an Indigenous environmental justice lens using document and archival research in the context of the treaty making process between Fort Albany FN and the British Crown, and the establishment of reserves. In brief, procedural issues were noted, as there was no transparency in reserve choice at the time of signing the treaty, and during the actual surveying of the reserve boundaries with certain types of land being excluded from reserve locations, unbeknownst to the FNs peoples. The Cree were also misled into believing that they would retain access to their whole traditional homeland―and not be confined to reserve land―the Cree believed that they only agreed to share the land. Historically, the Cree harmonized with the seasons and would not be residing in the Albany River floodplain during river freeze-up and during river break-up―adaptive behaviour to avoid flooding. Harmonizing with the environment had allowed the mobile Cree to live successfully with the annual flooding of the Albany River for millennia, until being forced to live permanently on reserve land by the colonial government. Nonetheless, the Cree still sustain their cultural worldview acknowledging the Cree cycle of life. The way forward for Fort Albany First Nation will be either relocation to high ground or trying to tame nature by reinforcing the existing dikes—or some novel combination of both based on two worldviews. Full article

We investigate the optical response properties of an atom-assisted spinning optomechanical system, in which a spinning optical resonator is coupled simultaneously to a two-level atomic ensemble and a mechanical resonator driven by a weak pump field. Remarkably, we demonstrate that by simply reversing the rotation direction, the system can be switched between a low-absorption electromagnetic and optomechanically induced transparency state and a high-absorption state, constituting a form of non-reciprocal optical control at the quantum level. Furthermore, by tuning the phase difference between the mechanical pump and the probe field, direction-dependent switching between absorption and gain is achieved. These non-reciprocal effects originate from the Sagnac-induced frequency shift in the optical mode, which leads to distinct optomechanical and atom–cavity couplings for opposite spinning directions. We also show that the absorption spectrum can be modulated by the angular velocity and the atomic number. Our results indicate that the optical properties of the hybrid system can be manipulated via the angular velocity, phase difference, and atom number, with potential applications in chiral photonic communications. Full article

Scientific evidence is fundamental to uncovering insights about health, including food and nutritional claims. Substantiating such claims requires robust scientific procedures that often include clinical studies, biochemical analyses, and the examination of multiple forms of data. The growing capabilities of artificial intelligence (AI) and large language models (LLMs) present new opportunities for analyzing food–health relationships and supporting health claim validation. Yet, applying these technologies to the food and nutrition domain raises challenges that differ from those encountered in broader biomedical text mining (TM). In this perspective, we review key issues, including the complexity and heterogeneity of food-related data, the scarcity of food-specific language models and standardized resources, difficulties in interpreting nuanced and often contradictory evidence, and requirements for integrating AI tools into regulatory workflows. We compare modern LLM approaches with traditional TM methods and discuss how each may complement the other. Our position is that, despite their promise, current AI and LLM tools cannot yet reliably handle the subtleties of food–health evidence without substantial domain-specific refinement and human expert oversight. We advocate for hybrid approaches that combine the precision of established TM techniques with the analytical breadth of LLMs, supported by harmonized ontologies, multidimensional evaluation frameworks, and human-in-the-loop validation, particularly in regulatory contexts. We also highlight the importance of public education, transparent communication standards, and coordinated cross-disciplinary efforts to ensure these technologies serve broader goals of food safety, consumer trust, and global health. Full article

Machine learning methods are increasingly applied in daylight performance assessment due to their ability to model complex nonlinear relationships within large datasets while offering substantially faster predictions than conventional simulation workflows. Within this framework, deep learning architectures provide enhanced representational capability for capturing spatial and geometric dependencies. However, existing approaches often lack seamless integration with parametric design environments and offer limited interpretability regarding the influence of design parameters. This paper presents DayANN (Daylight Artificial Neural Network), a feedforward deep neural network developed within a structured Grasshopper-to-machine learning workflow for analyzing daylight performance in a parametrically defined office space. The method employs Climate Studio for Grasshopper to generate 288 simulation scenarios, forming the training dataset for the predictive model. The proposed framework enables automated data transfer, model training, and performance feedback within an iterative design–evaluation loop. In addition to predictive accuracy, SHAP-based interpretability is incorporated to quantify the contribution of individual daylighting parameters. The model achieved high accuracy, with R² values of 0.988 for Useful Daylight Illuminance (UDI) and 0.947 for Daylight Factor (DF), demonstrating that DayANN serves as a computationally efficient, transparent surrogate model suitable for early-stage architectural decision-making. Full article

Large-scale post-marketing drug safety data from spontaneous reporting systems offer new opportunities to explore adverse drug events (ADEs). However, these datasets often contain high rates of missing and incomplete data, undermining the reliability and interpretability of pharmacovigilance analyses. Effective management of these data quality issues requires interactive tools to explore patterns of missingness across multiple dimensions. We present IRVINE (Interactive Visualization for Spontaneous Reporting Systems Databases Missing Values), an interactive visualization system designed to explore and compare missing data in spontaneous reporting systems. IRVINE integrates multiple coordinated components—including a global overview, detailed attribute-level breakdowns, a temporal analysis interface, and a cross-database comparison environment—allowing users to fluidly transition between global summaries and fine-grained diagnostic views. The system supports dynamic filtering, drill-down exploration, and interactive temporal analysis to examine changes in data completeness over time and across categories. Through three usage scenarios and a user study, we demonstrate how IRVINE supports effective exploration of reporting completeness. Results indicate that users perceived the system as easy to use and effective for identifying missingness patterns, with particular strengths in comparative and detail-level analysis. This work lays a foundation for improved transparency, interpretability, and data quality assessment in large-scale pharmacovigilance systems. Full article

The ability to assess molecular binding kinetics in real time is critical for advancing our understanding of molecular interactions in biochemical and biotechnological systems. This work presents a novel optical tweezer (OT)-based method to monitor molecular affinity in real time, focusing on the high-affinity streptavidin–biotin system as a model. Transparent poly(methyl methacrylate) (PMMA) microparticles functionalized with streptavidin were trapped before, during, and after binding with biotinylated bovine serum albumin (biotin–BSA), enabling the analysis of forward-scattered signals to detect nanoscale changes in particle size. By applying the Power Spectral Density method, the friction coefficient of individual particles was calculated, allowing for real-time tracking of binding dynamics and the estimation of the association rate constant ($k_{\mathrm{on}}\approx {10}^6{\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$). These results are consistent with literature values and demonstrate the potential of this OT-based approach for non-invasive, label-free detection of molecular interactions. Compared to existing techniques, such as atomic force microscopy and cantilever-based sensors, this method offers significant advantages, including real-time monitoring, adaptability to different bioaffinity systems, and compatibility with miniaturized setups. This work establishes a foundation for using OT-based tools to monitor high-affinity molecular interactions in real time. While demonstrated here using biotinylated BSA as a model ligand, future studies will explore the method’s applicability to smaller ligands and more subtle surface modifications. Full article

Modern supply chains are increasingly expected to meet ambitious sustainability targets, yet they often suffer from limited visibility into upstream relationships, environmental risks, and ethical sourcing practices. This paper presents an artificial intelligence (AI)-based approach for supporting sustainability-oriented decision-making in supply chains through knowledge graph completion and link prediction. We construct a multi-relational supply chain knowledge graph that captures heterogeneous entities and relationships, including suppliers, products, certifications, and locations, and apply graph neural networks to infer missing links and sustainability-related attributes. By enabling reasoning over incomplete and sparse data, the proposed approach supports feasibility-oriented decisions, such as identifying alternative supplier relationships and assessing sustainability alignment across multi-tier networks. Building on recent advances in knowledge graph reasoning and heterogeneous graph learning, the framework integrates relational structure with inductive learning to provide interpretable recommendations under uncertainty. The approach is evaluated on two real-world supply chain datasets, demonstrating its applicability in complex, data-sparse settings. The results indicate that graph-based AI can provide a practical foundation for transparent and sustainability-aware supply chain decision support. Full article

This study proposes a hierarchical signal-to-policy learning framework for risk-aware portfolio optimization that integrates model-based return forecasting, explainable machine learning, and deep reinforcement learning (DRL) within a unified architecture. In the first stage, next-period returns are estimated using gradient-boosted tree models, and SHAP-based feature attributions are extracted to provide transparent, factor-level explanations of the predictive signals. In the second stage, a Proximal Policy Optimization (PPO) agent incorporates both predictive forecasts and explanatory signals into its state representation and learns dynamic allocation policies under a mean–CVaR reward function that explicitly penalizes tail risk while controlling trading frictions. By separating signal extraction from policy learning, the proposed architecture allows the use of economically interpretable predictive signals to incorporate into the policy’s state representation while preserving the flexibility and adaptability of reinforcement learning. Empirical evaluations on U.S. sector ETFs and Dow Jones Industrial Average constituents show that the hierarchical framework delivers higher and stable out-of-sample risk-adjusted returns relative to both a single-layer DRL agent trained solely on technical indicators, a mean–CVaR optimized portfolio using the same parameters used in the proposed hierarchical model and standard equal weight as well as index-based benchmarks. These results demonstrate that integrating explainable predictive signals with risk-sensitive reinforcement learning improves the robustness and stability of data-driven portfolio strategies. Full article

Small-and-medium-sized enterprises (SMEs) increasingly depend on business partnerships to access markets and scale operations, yet they often face trust barriers during contract formation due to the complexity of the verification of their cybersecurity posture and compliance status by their partners. This problem is intensified by rising regulatory expectations, notably the EU Cyber Resilience Act (CRA), which many SMEs struggle to interpret and operationalize under constraints of budget, skills, and fragmented responsibilities. This study adopts a Design Science Research approach to blueprint and evaluate a lightweight mapping framework that links commonly implemented security controls to CRA requirements and to widely recognized benchmarks (ISO/IEC 27001 and CIS). Grounded in Institutional Theory and Socio-Technical Systems Theory, the artefact translates regulatory obligations into actionable, evidence-backed controls and produces partner-facing outputs that support transparency in negotiations and service level agreements. The framework is iteratively co-created with a multidisciplinary expert community. Expected contributions include a practical mechanism for making cybersecurity maturity visible, accelerating partnership formation, and enabling sustainable interorganizational relationships while remaining feasible for resource-constrained SMEs. Full article