Search Results (34,046)

Background/Objectives: Dental diseases represent a great problem for oral health care, and early diagnosis is essential to reduce the risk of complications. Panoramic radiographs provide a detailed perspective of dental structures that is suitable for automated diagnostic methods. This paper aims to investigate the use of an advanced deep learning (DL) model for the multiclass classification of diseases at the sub-diagnosis level using panoramic radiographs to resolve the inconsistencies and skewed classes in the dataset. Methods: To classify and test the models, rich data of 10,580 high-quality panoramic radiographs, initially annotated in 93 classes and subsequently improved to 35 consolidated classes, was used. We applied extensive preprocessing techniques like class consolidation, mislabeled entry correction, redundancy removal and augmentation to reduce the ratio of class imbalance from 2560:1 to 61:1. Five modern convolutional neural network (CNN) architectures—InceptionV3, EfficientNetV2, DenseNet121, ResNet50, and VGG16—were assessed with respect to five metrics: accuracy, mean average precision (mAP), precision, recall, and F1-score. Results: InceptionV3 achieved the best performance with a 97.51% accuracy rate and a mAP of 96.61%, thus confirming its superior ability for diagnosing a wide range of dental conditions. The EfficientNetV2 and DenseNet121 models achieved accuracies of 97.04% and 96.70%, respectively, indicating strong classification performance. ResNet50 and VGG16 also yielded competitive accuracy values comparable to these models. Conclusions: Overall, the results show that deep learning models are successful in dental disease classification, especially the model with the highest accuracy, InceptionV3. New insights and clinical applications will be realized from a further study into dataset expansion, ensemble learning strategies, and the application of explainable artificial intelligence techniques. The findings provide a starting point for implementing automated diagnostic systems for dental diagnosis with greater efficiency, accuracy, and clinical utility in the deployment of oral healthcare. Full article

Reliable knowledge of power transmission tower locations is fundamental for large-scale inspection and asset management in modern power grids. However, in satellite and aerial remote sensing imagery, towers typically appear as small, slender structures embedded in cluttered backgrounds, which leads to frequent missed and false detections. To address this challenge, we propose SCOPE-YOLO, an integrated super-resolution-plus-detection framework tailored for scalable transmission and distribution tower monitoring. In the first stage, low-resolution image patches are enhanced by a Real-ESRGAN ×4 super-resolution frontend, which restores high-frequency lattice details and sharpens tower boundaries. The reconstructed images are then processed by SCOPE-YOLO, a YOLOv11-based detector that incorporates a Cross-Scale Feature Aggregation (CFA) module, a Gather–Distribute (GD) routing mechanism, and a high-resolution P2 detection head, together with SAT and layered inference strategies to strengthen small-object perception under complex backgrounds. Experiments on the public SRSPTD dataset demonstrate that SCOPE-YOLO improves F1 score by 0.051 and raises mAP@0.5 by 10.2 percentage points over the YOLOv11-s baseline, while maintaining a compact model size. Compared with a broad set of state-of-the-art detectors, SCOPE-YOLO achieves the best overall performance, reaching 82.8% mAP@0.5 for power tower detection. Cross-domain evaluation on the GZ-PTD test set further confirms the effectiveness of the super-resolution–detection pipeline: Real-ESRGAN×4@2048 + SCOPE-YOLO increases Recall from 0.8621 to 0.9278 and mAP@0.5 from 0.8365 to 0.9132 relative to the low-resolution baseline, substantially reducing missed detections of small and weak tower targets in real-world scenes. Full article

The rapid global spread of antimicrobial resistance (AMR) has significantly reduced the effectiveness of many modern antibiotics, creating an urgent need for alternative therapeutic strategies. One promising approach is the revival and repurposing of older antimicrobial agents whose clinical potential was previously limited by toxicity concerns, pharmacokinetic challenges, or the availability of newer drugs. Recent advances in drug delivery, dosing optimization, and antimicrobial stewardship have renewed interest in these compounds as viable options for the treatment of multidrug-resistant infections. The aim of this review is to provide a comparative, clinically oriented analysis of selected “old” antibiotics, fosfomycin, colistin, streptomycin, and vancomycin, with emphasis on their current therapeutic roles, pharmacokinetic/pharmacodynamic (PK/PD) targets, toxicity mitigation strategies, resistance mechanisms, and evidence supporting combination therapies and alternative routes of administration. This narrative review was conducted using a structured PubMed search and manual reference screening, focusing on clinical, PK/PD, and translational studies relevant to the contemporary use of legacy antibiotics. The review summarises current evidence on the re-emerging clinical applications of these agents, each discussed in the context of historical use, mechanism of action, resistance patterns, and newly identified indications. Attention is given to novel formulations, combination strategies, and alternative routes of administration that enhance efficacy while limiting toxicity, including applications in biofilm-associated infections. Overall, strategic repurposing of older antibiotics represents a valuable complementary approach in the fight against AMR and may extend the therapeutic lifespan of existing agents in an era of limited antibiotic innovation. Full article

Active athletes represent a specific target for learning and development. Their schedules, including training sessions and competitions, leave little time for education. However, athletes still need skills beyond sports to ensure they are prepared for future employment. Our study approaches this issue by identifying appropriate settings for athletes’ learning and development. (1) Based on the background of current athletes’ education, it addresses the gap of not enough attention being paid to transferable practices from corporate attitudes to learning and development. (2) The study’s methodology primarily uses the case study concept because this conveys the video content we created for the athletes’ learning and development. This is combined with the method of content analysis of selected examples from corporate learning and development and the design thinking workshop, with the engagement of important stakeholder groups: athletes (2 participants), lecturers (2 participants), and representatives of sports organizations (1 participant). The other 9 workshop participants were master’s students in a managerial study programme because of their age similarities with the current athletes and the applicability of the courses they were studying to athletes’ education. (3) The designed process was created as a digital twin using haptic artefacts and the S2M technology (version 1.0) within the OMiLAB platform (version 1.6). Our results show that video content tailored to the athletes’ constraints is a viable solution that improves their career prospects. (4) The study’s practical implications are supported by the expert validation of the model provided by the inside of the large sports organizations’ management. Full article

To meet the demand for high efficiency in modern broadband communication systems, this paper presents a novel continuous-mode Doherty power amplifier design method based on integrated high-order filter prototypes. By deeply merging the filter structure with the output matching network, broadband impedance transformation and harmonic suppression are simultaneously achieved within the 1.6–2.2 GHz frequency range. This approach resolves the bandwidth limitations and efficiency degradation caused by the conventional separation of matching and harmonic control stages. Using a CGH40010F GaN transistor, the impedance space was determined through load-pull analysis, and the design flexibility was enhanced by applying continuous Class-F mode theory. The implemented amplifier demonstrates a saturated efficiency of 68–72%, a 6 dB back-off efficiency of 58.9–64.9%, a saturated output power exceeding 45 dBm, an in-band gain greater than 11.2 dB, and a return loss better than −15 dB. The proposed method offers an effective solution for the design of high-performance broadband power amplifiers. Full article

Modern geo-information platforms commonly adopt multi-window map interfaces that integrate heterogeneous data, such as dynamic maps and live camera feeds. These interfaces impose high cognitive load and slow spatial event detection. Operators must rapidly locate the source of visual alarms, a task often leading to delays under high visual workload. To address this challenge, this study investigated whether spatialized auditory cues can improve alarm localization in such complex monitoring interfaces. A controlled experiment with 24 participants used a within-subjects design to test factors of auditory spatial cueing (none, binaural, monaural), display dynamics (dynamic, static), and interface complexity (4, 8, 12 panes). Behavioral and eye-tracking data measured detection accuracy, efficiency, and gaze patterns. Results showed that dynamic displays and high interface complexity impaired performance, indicating increased cognitive load. In contrast, monaural lateralized auditory alarms substantially improved detection efficiency and mitigated visual overload. Interaction analyses revealed that binaural cues reduced the performance costs of dynamic displays, whereas monaural cues compensated for high-density layouts. These findings demonstrate that spatialized auditory alarms effectively support spatiotemporal situational awareness and improve operator performance in high-load geo-surveillance systems. The study offers empirical and practical implications for designing cognitively ergonomic, multimodal interfaces that move beyond purely visual alarm designs. Full article

Allosteric transcription factor (aTF)-based in vitro biosensors constitute a class of detection tools formed by the functional coupling of the ligand-binding domain of aTFs with a reporter system. Owing to advantages such as high specificity and sensitivity, these biosensors have emerged as a research hotspot in the field of modern food contaminant monitoring. Our work centers on the core aspect of engineering design and systematically elaborates on the modular design strategies for aTF-based in vitro biosensors, with a focus on the design principles of the molecular recognition system, signal amplification strategies, signal output systems, and sensing systems. Furthermore, the article summarizes the advances in the application of aTF biosensors for detecting various typical food contaminants and analyzes their performance advantages. Finally, in light of existing technical limitations, it prospectively discusses future directions for enhancing specificity, improving stability, and promoting commercial applications, aiming to provide a theoretical reference and application guidance for transitioning this technology from laboratory platforms to on-site real-time monitoring. Full article

The Turkish economy has been affected by recurring populist cycles and resultant economic crises, which have, in turn, unfavorably influenced the growth performance of the country. Inspired by the Turkish experience, this study attempts to investigate the effects of changes in exchange rate on the growth performance of the Turkish economy by using the production function framework. The data is sourced from the World Development Indicators and Penn World Table. Modern time series techniques are utilized to estimate the production function. Our findings reveal that there is a long-term but unfavorable relationship between changes in the exchange rate and economic growth in Turkey over the 1980–2019 period. Beyond its macroeconomic implications, the findings highlight that persistent exchange rate instability undermines macroeconomic sustainability by distorting the price mechanism, weakening investment incentives, and reducing long-term productive capacity. In this context, exchange rate stability emerges as a critical prerequisite for achieving sustainable economic growth in emerging economies such as Turkey. Full article

This essay explores the metaphysical and philosophical implications of Artificial Intelligence (AI) and Machine Learning (ML) through the intersecting insights of René Guénon (ʿAbd al-Wāḥid Yaḥiā), Martin Heidegger, and Ibn al-ʿArabī. It argues that modern AI systems, particularly in their statistical and data-centric forms, are not merely instrumental tools but expressions of a deeper metaphysical worldview-one rooted in quantification, abstraction, and utility. Guénon’s critique of the “reign of quantity” and Heidegger’s notion of Enframing (Gestell) converge in diagnosing the loss of qualitative and sacred dimensions in modern life. While Heidegger’s phenomenology provides a powerful immanent critique of technological reductionism from within the Western philosophical tradition, Guénon’s metaphysical traditionalism articulates a diagnosis of modernity that resonates with Islamic metaphysics, especially as articulated by Ibn al-ʿArabī. The essay includes Heidegger in the argument as a representative of a critique of modern technology issuing from the Western tradition itself, and by emphasizing his shared concerns with Guénon, whose metaphysics resonates with Ibn al-ʿArabī’s metaphysics. Through a comparative metaphysical framework, this paper proposes an Islamic response to AI that avoids both technophilia and technophobia, insisting instead on a spiritually grounded ethic of technology that preserves human’s dignity and mission. Methodologically, the essay restores a prior order often inverted in contemporary AI ethics: ontology (what AI is) grounds epistemology (what it can know), and only then can ethical evaluation be coherent. Full article

Situated within the field of modern Chinese political history, this study investigates the Late Qing New Policies (1901–1911) as a pivotal transition from a traditional tributary empire to a modern multi-ethnic nation-state. A critical limitation in current scholarship is the tendency to reduce these reforms to mere expedients for dynastic preservation, thereby overlooking the complex mechanisms by which they fundamentally reconstructed national identity and interethnic power structures amidst the “triple crisis” of territory, sovereignty, and nationality. To address this, the article employs a comprehensive historical analysis to explore how institutional restructuring in administration, military, and ideology catalyzed the transformation from imperial autocracy toward a “responsible government” framework. The research is distinguished by its innovative application of Anthony D. Smith’s theories of “ethnic” versus “civic” nationalism to deconstruct the “myth-symbol complex” of the Chinese nation, bridging the theoretical divide between the “New Qing History” paradigm and empirical modernization narratives. Findings demonstrate that while the Manchu leadership aimed to secure formal primacy, the practical implementation of reforms engendered a de facto Han-supported power structure, compelling the reconceptualization of the state as a “multi-ethnic constitutional monarchy” and establishing the institutional logic for the “Five Races Under One Union” model. Consequently, this study offers significant academic value by redefining the New Policies as the foundational phase of modern China, providing a crucial theoretical framework for understanding the continuity of China’s multi-ethnic statehood and national identity beyond the dynastic collapse. Full article

Short-Term Residential Load Forecasting (STRLF) is a core task in smart grid dispatching and energy management, and its accuracy directly affects the economy and stability of power systems. Current mainstream methods still have limitations in addressing issues such as complex temporal patterns, strong stochasticity of load data, and insufficient model interpretability. To this end, this paper proposes an explainable and efficient forecasting framework named KAN+Transformer, which integrates Kolmogorov–Arnold Networks (KAN) with Transformers. The framework achieves performance breakthroughs through three innovative designs: constructing a Reversible Mixture of KAN Experts (RMoK) layer, which optimizes expert weight allocation using a load-balancing loss to enhance feature extraction capability while preserving model interpretability; designing an attention-guided cascading mechanism to dynamically fuse the local temporal patterns extracted by KAN with the global dependencies captured by the Transformer; and introducing a multi-objective loss function to explicitly model the periodicity and trend characteristics of load data. Experiments on four power benchmark datasets show that KAN+Transformer significantly outperforms advanced models such as Autoformer and Informer; ablation studies confirm that the KAN module and the specialized loss function bring accuracy improvements of 7.2% and 4.8%, respectively; visualization analysis further verifies the model’s decision-making interpretability through weight-feature correlation, providing a new paradigm for high-precision and explainable load forecasting in smart grids. Collectively, the results demonstrate our model’s superior capability in representing complex residential load dynamics and capturing both transient and stable consumption behaviors. By enabling more accurate, interpretable, and computationally efficient short-term load forecasting, the proposed KAN+Transformer framework provides effective support for demand-side management, renewable energy integration, and intelligent grid operation. As such, it contributes to improving energy utilization efficiency and enhancing the sustainability and resilience of modern power systems. Full article

In modern marine seismic exploration, ocean bottom node (OBN) acquisition systems are increasingly valued for their flexibility in deep-water complex structural surveys. However, the high operational costs associated with OBN systems often lead to spatially sparse sampling, which adversely affects the fidelity of wavefield reconstruction. To overcome these limitations, hybrid deep learning frameworks that integrate physics-driven and data-driven approaches show significant potential for interpolating OBN four-component (4C) seismic data. The proposed frequency-domain residual-attention U-Net (ResAtt-Unet) architecture systematically exploits the inherent physical correlations among 4C data to improve interpolation performance. Specifically, an innovative dual-branch dual-channel network topology is designed to process OBN 4C data by grouping them into complementary P–Z (hydrophone–vertical geophone) and X–Y (horizontal geophone) pairs. A synchronized joint training strategy is employed to optimize parameters across both branches. Comprehensive evaluations demonstrate that the ResAtt-Unet achieves superior performance in component-wise interpolation, particularly in preserving signal fidelity and maintaining frequency-domain characteristics across all seismic components. Future work should focus on expanding the training dataset to include diverse geological scenarios and incorporating domain-specific physical constraints to improve model generalizability. These advancements will support robust seismic interpretation in challenging ocean-bottom environments characterized by complex velocity variations and irregular illumination. Full article

Polynucleotides purified through High Purification Technology (PN HPT), combined with hyaluronic acid (HA), represent a novel injectable strategy to improve skin quality in aesthetic medicine. This real-world data collection aimed to evaluate the safety and performance of PN HPT-based treatments across multiple facial and body areas. Data were collected through a post-market clinical follow-up survey, analysing 218 questionnaires completed after 654 intradermal infiltrations performed on the face (e.g., forehead, perioral lines, crow’s feet), neck, hands, and décolleté. Aesthetic outcomes were assessed using clinician- and patient-reported Global Aesthetic Improvement Scale and Global Clinical Improvement Scale scores. Safety and satisfaction were evaluated through adverse event reporting and a patient-completed Likert scale. Across all treated areas, consistent improvement was observed. For the face, 39% of cases achieved “marked improvement” or “excellent result” on the GCI-S, with 48% rated as “much improved” or “very much improved” on the GAIS. Similar outcomes were reported for the neck (41% and 57%), hands (31% and 41%), and décolleté (43% and 55%). Patient satisfaction was high, with over 90% expressing general satisfaction and willingness to repeat the treatment. No serious or unexpected adverse events occurred. These findings suggest that PN HPT and HA injectables may offer a high level of patient satisfaction, observable improvements in skin quality, and a favorable safety profile in the modern aesthetic practice. Full article

With artificial intelligence (AI) rapidly increasing in popularity and presence in everyday life, new applications utilizing AI are being explored across virtually all domains, from banking and healthcare to cybersecurity to generative AI for images, voice, and video content creation. With that trend comes an inherent need for increased AI capabilities. One cornerstone of AI applications is the ability of generative AI to consume documents and utilize their content to answer questions, generate new content, correlate it with other data sources, and more. No longer constrained to text alone, we now leverage multimodal AI models to help us understand visual elements within documents, such as images, tables, figures, and charts. Within this realm, capabilities have expanded exponentially from traditional Optical Character Recognition (OCR) approaches towards increasingly utilizing complex AI models for visual content analysis and understanding. Modern approaches, especially those leveraging AI, are now focusing on interpreting more complex diagrams such as flowcharts, block diagrams, Unified Modeling Language (UML) diagrams, electrical schematics, and timing diagrams. These diagram types combine text, symbols, and structured layout, making them challenging to parse and comprehend using conventional techniques. This paper presents a historical analysis and comprehensive survey of scientific literature exploring this domain of visual understanding of complex technical illustrations and diagrams. We explore the use of deep learning models, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformer-based architectures. These models, along with OCR, enable the extraction of both textual and structural information from visually complex sources. Despite these advancements, numerous challenges remain, however. These range from hallucinations, where the content extraction system produces outputs not grounded in the source image, which leads to misinterpretations, to a lack of contextual understanding of diagrammatic elements, such as arrows, grouping, and spatial hierarchy. This survey focuses on five key diagram types: flowcharts, block diagrams, UML diagrams, electrical schematics, and timing diagrams. It evaluates the effectiveness, limitations, and practical solutions—both traditional and AI-driven—that aim to enable the extraction of accurate and meaningful information from complex diagrams in a way that is trustworthy and suitable for real-world, high-accuracy AI applications. This survey reveals that virtually all approaches struggle with accurately extracting technical diagram information. It also illustrates a path forward. Pursuing research to further improve their accuracy is crucial for supporting and enabling various applications, including complex document question answering and Retrieval Augmented Generation (RAG), document-driven AI agents, accessibility applications, and automation. Full article

Mathematical and computational models, which have been successfully used in various fields of biology, are particularly relevant in studies on the origin of life, where wet experiments have not yet been able to obtain fully “living” entities from abiotic materials. This paper investigates mathematical and computational models of interacting polymers in prebiotic environments to understand how molecular replication and protocell reproduction could emerge. This study builds on the Binary Polymer Model (K-BPM), in which polymers are represented as binary strings that undergo catalyzed condensation and cleavage reactions, by introducing a biologically relevant variant (C-BPM), where catalytic activity depends on polymer structure. The model is analyzed with respect to the formation of autocatalytic networks, formalized as Reflexive Autocatalytic Food-generated (RAF) sets, embedded in a protocell in order to simulate their dynamics. The results show clear differences between K-BPM and C-BPM models. They also show that the existence of a RAF does not guarantee the survival of a population of protocells, although it may be possible when only a subset of the existing species partakes in the RAF, thus suggesting that small autocatalytic networks may have preceded the larger networks found in modern life. Full article