Search Results (54,480)

The development of sustainable and efficient sample-preparation strategies remains a key challenge in analytical workflows for plant-derived bioactive compounds. In this study, choline chloride (ChCl)-based natural deep eutectic solvents (NADES) were systematically investigated as green extraction media and combined with chemical and enzyme assays to identify the bioactive effect of phenolic compounds from Bellevalia dubia, an understudied Mediterranean plant. A design of experiments approach was applied to evaluate the influence of hydrogen bond donors (lactic, malic, and citric acid) and their molar ratios on extraction performance. Among the tested systems, the ChCl–malic acid (MA) mixture at a molar ratio of 1:2.5 (precisely 1:2.47) was identified as the best overall solution according to the desirability approach, which integrated both phytochemical and enzyme-inhibitory responses. This system also showed strong enzyme-inhibitory effects compared to conventional extractants (water and 80% methanol). To identify the extract composition, a suspect screening workflow was applied based on ultra-high-performance liquid chromatography hybrid quadrupole Orbitrap mass spectrometry (UHPLC–q-Orbitrap-MS). Twenty-one compounds were detected and annotated, mainly belonging to the flavonoid class. Principal component analysis (PCA) showed that ChCl-MA extracts were discriminated, particularly at molar ratio 1:2.5, from the rest, indicating a distinct chemical and bioactivity profile. PCA further suggested a possible association of the ChCl-MA 1:2.5 system with vitexin and AChE inhibition, which requires further investigation. Overall, the proposed NADES-based approach offers a tunable, promising, and potentially more sustainable alternative for the extraction of bioactive compounds from plant-based sources. Full article

Background/Objectives: Musculoskeletal pain is a common condition that negatively affects quality of life and functional capacity. Integrative nursing emphasizes patient-centered and non-pharmacological approaches; however, variability in clinical decision-making persists due to reliance on clinician experience. This study aimed to design, retrospectively train, and evaluate the preliminary feasibility of a machine learning-based clinical decision support framework for personalized pain management within an integrative nursing approach in a Traditional and Complementary Medicine setting. Methods: This study included a retrospective cohort of 487 patients (mean age: 48.2 ± 14.7 years; 63.0% women) for machine learning model development and evaluation. The classification model predicted binary clinical improvement, defined as a ≥3-point reduction in VAS pain score, whereas the regression model predicted post-treatment VAS score. The resulting models were incorporated into the GETAIA prototype to generate modality-specific predicted post-treatment VAS scores and probabilities of clinically meaningful improvement for the three available T&CM modalities; these predicted profiles were compared to provide a model-derived treatment recommendation for clinical review. In addition, a prospective, non-randomized pilot involving 42 patients (21 AI-assisted and 21 standard care) was conducted to assess preliminary point-of-care workflow feasibility. Results: Significant reductions in pain intensity were observed across all treatment groups (p < 0.001). Clinically meaningful improvement rates were 76.2% for cupping therapy, 68.3% for acupuncture, and 58.8% for mesotherapy (p = 0.007). The prototype demonstrated 78.3% concordance with historical clinician treatment decisions, which was interpreted as an internal model-fidelity measure reflecting the reproduction of historical clinical decision patterns rather than evidence of improved decision-making, clinical effectiveness, or external validity. During a preliminary prospective feasibility implementation, clinically favorable numerical trends were observed in the AI-assisted group compared with standard care, including clinically meaningful improvement in 85.7% (18/21) versus 71.4% (15/21) of patients (p = 0.449) and mean VAS reductions of 3.9 ± 1.2 versus 3.2 ± 1.4 (p = 0.092). These exploratory differences were not statistically significant and were not intended to establish comparative clinical effectiveness. On the held-out test set, the primary XGBoost classification model achieved an accuracy of 82.6% and an AUC-ROC of 0.87 (95% CI: 0.82–0.92), while the regression model predicting post-treatment VAS achieved an MAE of 0.87. Key predictors of improvement included diagnosis, baseline pain intensity, and age. Conclusions: The findings support the preliminary development and evaluation of a machine learning-based decision-support framework and GETAIA prototype for individualized pain management within an integrative nursing context. The retrospective analysis demonstrated predictive performance of the evaluated models, while the small, non-randomized prospective pilot provided preliminary information on point-of-care workflow integration. These findings do not establish the clinical effectiveness, superiority, routine-care readiness, or external validity of a fully implemented clinical decision support system. Larger, prospective, multicenter studies are required to evaluate clinical effectiveness, generalizability, usability, and implementation readiness. Full article

This paper provides the design and simulation of a lightweight cryptographic protocol on smart house IoT devices using ChaCha20 and Ascon-AEAD128. The protocol, implemented in Python 3.14.6 and tested on a virtual ESP32 platform using Wokwi and CloudAMQP, uses stream cipher encryption with authenticated message tagging to provide confidentiality and integrity. Two conditions, normal and tampered transmission, were experimented to confirm tag validation and successful decryption. Findings affirmed sound detection of tampering and unauthorized access prevention, proving usefulness of current AEAD ciphers on limited devices. The protocol highlights cryptographic systems that prioritize computationally efficiency and robustness, which is essential in smart homes that have limited power and memory. The hybrid design provides confidentiality and authenticity using a minimal overhead by utilizing ChaCha20 to provide lightweight encryption and Ascon-AEAD128 as authentication. The resilience to the replay and modification attacks was demonstrated in experiments based on message queues and injected packet modifications simulating real-world conditions. Even though benchmarking of hardware was not carried out, the simulated values reveal stability and flexibility to use in low-power systems. This article emphasizes the necessity of authenticated encryption as default, which is consistent with the NIST standards and reflects the appropriateness of Ascon to new IoT security requirements. It also creates a reconfigurable structure that can be used in other highly constrained systems, such as healthcare monitoring and industrial IoT. The main contribution is the gap between theoretical cryptography and practical IoT security provided by the practical prototype. Further development will include tests on physical ESP32 modules and fine performance profiling, yet already, the current implementation proves a scalable, secure model of smart home IoT. Finally, this research demonstrates that the demand of reliable and low-power-based communication in resource constrained networks can be met efficiently without sacrificing device performance by means of lightweight cryptography. In addition to the cryptographic protocol design, this study integrates a machine learning-based intrusion detection layer trained on the Edge-IIoTset dataset, in which Logistic Regression, Decision Tree, Random Forest, and Gradient Boosting classifiers are evaluated to complement the encryption–authentication framework with anomaly-aware monitoring of network traffic. Full article

Near-duplicate diagram detection in electronic documents is challenging because diagram identity depends on graphical structure, spatial composition, and textual labels, while reused images may undergo compression, cropping, rotation, photometric changes, or perspective distortion. This study proposes a cascaded multimodal framework combining perceptual hashing, Siamese Residual Network with 18 layers (Siamese ResNet18), Distillation with No Labels, ver. 2 (DINOv2)visual representations, and a text-similarity classifier. A controlled benchmark was constructed from Artificial Intelligence 2D Diagram Dataset (AI2D) using Light, Medium, and Hard transformations, with source-grouped splitting by base_id to prevent leakage across training, validation, and test sets. Perceptual hashing achieved Area Under the Receiver Operating Characteristic Curve (ROC-AUC) = 0.7483, while Siamese ResNet18 increased ROC-AUC to 0.8599. DINOv2 provided the strongest visual performance, achieving Accuracy = 0.9933, F1-score = 0.9933, ROC-AUC = 0.9992, and Average Precision = 0.9994; F1-score remained 0.9901 for Hard transformations. Visual fusion increased ROC-AUC to 0.9995, and full multimodal fusion reached ROC-AUC = 0.9998. At an early-exit threshold of 0.95, 27.6% of pairs were resolved at the hashing level. These results support the coarse-to-fine design on the constructed AI2D-derived benchmark. A targeted hard-negative stress test revealed substantially higher false-positive rates under deliberately matched spatial layouts, with an overall False Positive Rate (FPR) of 0.48 for DINOv2 and 0.16 for full multimodal fusion. Generalization to naturally reused or redrawn diagrams, larger and more diverse hard-negative collections, and Optical Character Recognition (OCR)-derived text remains to be evaluated. Full article

The growing demand for large, lightweight, heat-resistant, and multifunctional aerospace structures has raised the requirements for metal matrix composites in terms of defect minimization, performance enhancement, and near-net-shape manufacturing. Additive friction stir deposition of composites enables feedstock delivery, reinforcement mixing, and layer-by-layer consolidation in a thermoplastic state below the melting point of the matrix, thereby mitigating porosity, hot cracking, elemental segregation, reinforcement degradation, and excessive interfacial reactions commonly encountered in fusion-based additive manufacturing. This review summarizes recent advances in the application of this technology to the fabrication of metal matrix composites, elucidates the mechanisms of material flow, interlayer bonding, microstructural evolution, and defect formation during deposition, and discusses the effects of tool design, process parameters, and reinforcement characteristics on interfacial bonding, microstructure control, and mechanical properties. Remaining challenges include the uniform delivery and quantitative control of reinforcements, characterization of interfacial bonding and load transfer, forming stability of complex components, and evaluation of in-service performance; accordingly, thermo-mechanical-flow multiphysics models, multisensor closed-loop control systems, and unified quality-assessment methods should be developed to promote the engineering application of large-scale, multimaterial graded aerospace components. Full article

Background: Neurosurgery depends heavily on implanted biomaterials, and the endovascular treatment of intracranial aneurysms (IAs) is the paradigmatic case, since coils, flow diverters and intrasaccular devices achieve durable occlusion only through intrasaccular thrombus organization and endothelial coverage of the neck, both governed by the local flow environment. Computational fluid dynamics (CFD) resolves that environment but is too slow and too solver-dependent for clinical or device design use. Graph and geometric deep learning operates natively on the unstructured meshes in which vascular anatomy and implanted scaffolds are represented and has been proposed as the technology that would close this gap. Methods: Four databases were searched up to 29 July 2026 for original studies developing or validating graph-based or geometric deep learning applied to IA geometry, with a hemodynamic or clinical outcome. Screening and extraction were performed in duplicate; risk of bias with PROBAST and artificial intelligence signaling items, reporting with TRIPOD+AI, and synthesis followed SWiM. Results: Twelve studies (2021 to 2026) were included: seven clinical (1965 aneurysms, plus 81 externally) and five computational (up to 984 geometries). Learned geometric representations discriminated rupture, growth and post-embolization recanalization better than morphological indices and then PHASES, reaching an area under the curve of 0.795 to 0.97 internally. Graph and point cloud surrogates reproduced hemodynamic fields with normalized errors of 2% to 5% in seconds rather than hours, and one transformer-based graph network predicted the extent and timing of intra-aneurysmal thrombus formation over a reactive surface. External validation was reached by two studies, and clinical utility was achieved in one; discrimination fell from 0.85 to 0.71 in one external test, and surrogate error rose from 4.1% to 19.1% on patient-derived anatomy. No model was trained on a device-laden geometry. Additionally, 5/7 clinical studies classified cross-sectional rupture status rather than prospectively predicting future rupture. Conclusions: Graph and geometric deep learning has substantially reduced the computational burden of hemodynamic analysis, but reliable generalization to unseen patient-specific anatomy remains a relevant challenge. Shared patient-specific benchmarks, prespecified external validation, reporting of calibration, and extension of training corpora to implanted anatomy are necessary next steps towards virtual evaluation of neurovascular biomaterials. Full article

This study evaluates the operational and performance outcomes of railway freight market liberalization across 28 European countries from 2013 to 2024, moving beyond simple operator counts to analyze key traffic, economic, and infrastructure metrics. To control for structural variation, we classified countries into small, medium, and large networks using K-means clustering based on railway route length. A Mixed-Design Repeated-Measures ANOVA was conducted to analyze temporal trends, network-scale impacts, and their interactions. The results reveal that while the total number of active operators rose steadily after 2018, other performance indicators showed an inverse relationship from 2020 onward. The network cluster had a highly significant main effect across all performance indicators, with the largest effect sizes for infrastructure access charges (partial η² = 0.638), freight train-kilometres (partial η² = 0.610), and net tonne-kilometres (partial η² = 0.585). The interaction between year and network cluster was also statistically significant, most notably for net tonne-kilometres (partial η² = 0.583). Underperformance was prominent in small networks, where market entry did not prevent long-term declines in traffic volume, whereas large networks achieved growth of over 25% in net tonne-kilometres. These quantitative findings show that counting active operators alone is an inadequate measure of liberalization success, as market performance and overall activity levels are strongly associated with network scale and capacity rather than market fragmentation. Full article

Building thermal performance has become a major concern as global warming intensifies and building energy demand is expected to rise by 30% by 2030. Bio-based construction materials offer a promising solution by improving thermal comfort and reducing operational energy use. This study presents a computational framework for optimising bio-based masonry block design through the integration of numerical simulation and data-driven modelling. A parametric thermal model was developed in COMSOL Multiphysics to simulate steady-state heat transfer through hollow masonry blocks with varying cavity geometries and material properties. A surrogate model-based approach was then used to generate a dataset of simulated block configurations, from which key thermal performance indicators, particularly heat flux, were extracted. These outputs were used to train a machine learning surrogate model capable of accurately predicting thermal performance across a wide design space without repeated finite-element simulations. The proposed workflow achieved an average heat-flux mismatch of 0.23%, reduced computational cost, and improved prediction accuracy by approximately 7%. In addition, the surrogate-assisted optimisation identified block geometries with substantially improved thermal performance compared with conventional reference configurations. The proposed methodology provides a scalable digital design framework for the development and evaluation of bio-based masonry materials. Future work will incorporate real-time sensor networks and Internet of Things (IoT) systems for the continuous monitoring and validation of masonry wall thermal performance. Full article

Accurate estimation of Behind-the-meter Photovoltaic (BTM PV) generation is essential for load forecasting and grid planning. Most distributed PV systems are installed behind customer meters, making their output unobservable. Disaggregating PV output from net load is therefore critical for improving distribution network observability. However, existing deep-learning-based disaggregation methods require large labeled datasets, and obtaining such data is costly. Under limited budgets, only a few users can be labeled, which constrains model performance. This paper proposes a user-level BTM PV disaggregation method based on active learning with adaptive weight updates, aiming to maximize model performance with minimal labeling cost. We design a multi-dimensional user value assessment system incorporating epistemic uncertainty, aleatoric uncertainty, and representativeness. In each iteration, the most informative users are selected for sub-meter installation. To dynamically optimize the selection strategy, we propose an adaptive weight update mechanism that adjusts the weights for the next round based on performance improvement gradients across dimensions. This closed-loop feedback enables the strategy to capture evolving model needs and prioritize users that yield the greatest performance gains. The proposed method is validated on the public Ausgrid dataset, and experimental results demonstrate its effectiveness under limited budgets. Full article

This paper presents a measurement-driven study of end-to-end latency in an indoor private standalone 5G network deployed at TU Wien IFT TEC-Lab. The testbed combines pico radio units, edge computing resources, and a local 5G core to form a campus-scale private network architecture designed for low-latency communication. To characterize network performance, TCP throughput and UDP one-way latency measurements were collected in a single-user, constant-bit-rate downlink setting at seven predefined indoor locations using iPerf-based tests at six traffic levels (1–500 Mbps), with five repetitions per condition. Based on these measurements, a compact parametric model was calibrated for this deployment to describe latency as a function of achieved bandwidth and location-dependent effects. The results show that latency remained low and relatively stable at low and medium traffic levels, generally staying below 20 ms between 1 and 200 Mbps, but increased more strongly as the operating point approached the practical throughput limit of the setup. The fitted model captured the overall latency trend with an in-sample MAE of 2.52 ms, an RMSE of 3.13 ms, and an $R^2$ of 0.842, while retaining comparable predictive performance under a trial-based test split ($R^2=0.835$) and leave-one-location-out validation ($R^2=0.818$). The compact model also achieved lower out-of-sample errors than the minimal M/M/1-type and polynomial-regression baselines under both validation schemes. Overall, the findings indicate that traffic load was the main driver of latency growth in the studied environment, while spatial effects remained measurable but secondary. The resulting formulation should be understood as an interpretable empirical model of end-to-end latency for one indoor private standalone 5G deployment under single-user, constant-bit-rate downlink conditions, rather than as a general latency model for private 5G networks. Within that scope, the model provides an interpretable description of the measured latency behavior and a basis for preliminary capacity assessment in this deployment. Its applicability to another private 5G network has not been established and would require a new measurement campaign, complete parameter re-estimation, and independent validation. Full article

The incorporation of Digital Twins (DT) into the healthcare industry marks a revolution in terms of adopting a more proactive and personalized approach towards patient care. The increasing complexity of technological tools employed within the healthcare environment leads to assessing the potential impacts of these digital models in collaboration with AI and IoT for increased efficiency and improved results. In this context, this study offers a systematic review of existing research regarding DTs in the field of healthcare, with specific consideration of hospital applications. An extensive literature search was performed within the Scopus database for peer-reviewed publications during the period from 2021 to 2026. Following a demanding screening process, 70 relevant articles were found that fulfilled the selection criteria. The review shows an emerging trend towards the application of AI-based Digital Twins in the real-time monitoring, predictive maintenance of medical devices, and planning surgeries. The paper analyses several key characteristics of healthcare DTs, including their design and architecture, and the benefits they generate. It also presents the challenges related to data integration and ethics surrounding virtual health models and recommendations for future research. In conclusion, this review demonstrates the revolutionary role of AI- and IoT-enabled Digital Twins in the transformation of hospitals’ infrastructures. This paper summarizes the latest developments and gaps in this field and offers a starting point for further research in this area. Full article

Based on a fixed-time extended state observer (FESO), this paper proposes a novel predefined performance control (PPC) method for robotic manipulators with false data injection (FDI) attacks and actuator faults. First, a mathematical model of robotic manipulators with parameter uncertainties and external disturbances is constructed. Then, to compensate for the FDI attacks and actuator faults, an extended state is introduced such that the FESO is designed. Moreover, to avoid the transformation from nonlinear constraints to unconstrained variables in PPC, the barrier Lyapunov function (BLF) is introduced. By adopting the novel PPC, tracking errors of the robotic manipulators are driven into a predefined region. Finally, simulations on a two-degree-of-freedom manipulator demonstrate that, compared with FTESO-based sliding mode control, the proposed method has shorter settling times and better tracking accuracy. Full article

The paper presents a low-cost agricultural monitoring platform based on embedded IoT sensing, LoRa mesh communication, and an Edge–Fog–Cloud architecture designed for real-time environmental monitoring in smart agriculture applications. The proposed system integrates distributed sensor nodes built around the Heltec Mesh Node T114 platform, combining an nRF52840 microcontroller with an SX1262 LoRa transceiver to support low-power, long-range communication in agricultural environments. Environmental data acquisition is performed using integrated BME280 temperature, humidity, and pressure sensors, capacitive soil moisture sensors, and TEMT6000 light sensors. A custom firmware architecture extending the Meshtastic framework was developed to support sensor integration, telemetry generation, packet forwarding, and energy-aware scheduling within the LoRa mesh network. The platform combines embedded telemetry modules, intermediate routing mechanisms, and cloud-based microservices responsible for data aggregation and processing. Experimental results showed stable radio-link conditions in the tested node-to-router configuration, with RSSI values around −55 dBm and SNR values of approximately 6–7 dB. Channel utilization varied between approximately 4% and 12%, while transmission airtime utilization remained below 5%. These results support the feasibility of the proposed sensing, communication, and data-processing architecture under the tested configuration. The architecture is designed to accommodate additional sensing and routing nodes, while multi-hop performance, long-term energy consumption, and network scalability remain to be evaluated in larger deployments. Full article

As a bio-inspired paradigm fully implementable in optics, deep reservoir computing (RC) has emerged as a powerful computational framework for efficient information processing. However, conventional deep RC architectures typically demand substantial hardware overhead to establish deep nonlinear mapping. To address this, we propose a high-speed, single-physical-hardware deep photonic RC architecture based on an optically pumped spin vertical-cavity surface-emitting laser (spin-VCSEL). At the architectural level, an all-optical deep RC structure is established within a single-physical-hardware platform by feedforward-injecting the right-circularly polarized response into the left-circularly polarized mode, fully exploiting the nonlinear dynamics without another reservoir laser. Algorithmically, a compression–decompression framework is integrated to mitigate the latency overhead induced by time-division multiplexing without compromising performance. Such state reconstruction enables a tenfold reduction in the required optical-domain time-division multiplexing (TDM) processing duration. Numerical simulations demonstrate that the proposed system exhibits superior precision compared to conventional setups, yielding a normalized mean square error of 0.0039 in Santa Fe time-series prediction, a symbol error rate of 0.003 in nonlinear channel equalization, and a linear memory capacity of 19.82. Cross-correlation analysis supports reliable information transfer between the two polarization modes. Ultimately, this hardware-software co-designed paradigm offers a compact, low-latency platform for optical neuromorphic processing. Full article

Multimodal emotion recognition systems based on electroencephalography (EEG) and eye tracking (ET) provide complementary information about neural and visual responses; however, practical deployment requires balancing classification performance and computational cost. This study investigates this trade-off by systematically evaluating 60 feature representation configurations generated from different EEG channel regions, frequency bands, feature types, and signal modalities. Support Vector Machine (SVM), Random Forest (RF), and Artificial Neural Network (ANN) classifiers were evaluated under subject-independent leave-one-subject-out (LOSO) cross-validation, resulting in 180 realizable classifier-feature configuration systems. Median Macro-F1 was used as the primary performance objective, while median classifier inference time was used as the computational-cost objective. A multi-stage selection strategy integrating Δ-based near-optimal filtering and Pareto dominance analysis was applied to identify performance-efficient systems. The highest median Macro-F1 (0.6474) was achieved by an ANN using temporal delta-band PSD features combined with ET information, with a median classifier inference time of 0.0057 s. This system remained the final selected system across Δ values of 0.01, 0.02, and 0.03. An ET-only ANN baseline achieved a median Macro-F1 of 0.6265, indicating a modest improvement when temporal delta-band EEG information was added. These findings demonstrate that feature representation, modality composition, classifier choice, and computational cost should be considered jointly when designing subject-independent multimodal emotion recognition systems. Full article