Search Results (194)

The proliferation of Edge-AI in personalized healthcare has driven significant demand for energy-efficient Systems-on-Chip (SoCs) capable of executing high-accuracy, real-time disease detection. Conventional accelerator architectures struggle to simultaneously accommodate disparate AI models. Specifically, memory-intensive Convolutional Neural Networks (CNNs) for high-fidelity feature extraction and event-driven Spiking Neural Networks (SNNs) for ultra-low-power, brain-inspired computation. To address this bottleneck, this paper presents a comprehensive survey of Reconfigurable Hybrid Neural Networks (RHNNs), an emerging paradigm that dynamically merges the strengths of CNNs and SNNs to meet the stringent resource constraints of biomedical edge devices. We establish a comprehensive taxonomy of existing RHNN architectures, categorizing them by hardware interconnection topologies, dataflow orchestration strategies, and internal structural adaptation mechanisms. Furthermore, we examine the integration of these hybrid accelerators within the open-source RISC-V processor ecosystem, evaluating how custom instruction set extensions optimize control efficiency and minimize energy overhead. The survey also analyzes commonly used datasets based on three major biomedical signal modalities, including electroencephalography (EEG), electrocardiography (ECG), and electromyography (EMG), in the context of processing systems for hardware accelerators. Finally, we highlight the open research challenges and outline future research directions to guide the development of next-generation biomedical intelligent systems. Full article

SrAl₂O₄: Eu²⁺, Dy³⁺ polymeric composites are the reference photoluminescent materials in passive signaling due to their high emission intensity and long luminescent persistence. But their real long-term outdoor durability is still poorly understood. This study analyzes a strontium aluminate polymer composite exposed to real weathering for 15 years to evaluate its degradation threshold, exceeding the frameworks of accelerated tests. Using FTIR, Raman, FE-SEM, colorimetry and phosphorescence decay, the weathered material was compared with its reference material. Results show that the luminescent composite retains its structural properties, its optical functionality improving its mechanical properties (increasing the flexural strength between 23 and 64% and microhardness between 136 and 164%), while suffering only a small yellowing, suggesting a longer service life than expected. Finally, it was established that the loss in the optical performance is not due to irreversible degradation of the polymer or pigment but is caused by the accumulation of surface dust and the formation of an opaque outer layer. The application of mechanical surface polishing removes this polluting layer, restoring optical transmittance and effectively recovering almost the original luminosity of the photoluminescent system. Full article

Quantifying the direction, strength and temporal clustering of cross-border financial risk contagion calls for methods that go beyond linear correlation. We suggest a two-layer framework that brings together transfer entropy and a multivariate Hawkes self-exciting point process on a time-varying, directed, weighted network. In the first layer, one-to-one transfer entropies of sovereign credit default swap spreads are estimated with a bias-corrected k nearest neighbour estimator, and this step detects nonlinear and directional information transfer between spreads. The second layer is a multivariate Hawkes process that models how extreme loss events arrive and mutually excite one another across countries, and it gives an excitation intensity matrix, encoding the way a tail event in one country raises the likelihood of an instantaneous hazard occurring in another. By merging these two layers, we obtain a composite, directed, weighted adjacency matrix in which the weights of the edges reflect both information flow and event clustering. We introduce a network-level contagion intensity index and split it into direct, indirect and feedback terms using the graph Laplacian spectrum. Von Neumann graph entropy together with the spectral gap ratio serve as entropy-based measures of the complexity and fragility of the evolving network. We validate the choice of Shannon-type entropy through a Tsallis q-sensitivity analysis, and we verify the nonlinear dependence structure of the data using BDS tests and maximal Lyapunov exponent estimates. Three empirical findings emerge from analysing 20 sovereign CDS markets from January 2015 to December 2025: (i) directional risk spillover signals derived based on transfer entropy are more timely than those derived from variance decomposition; (ii) the Hawkes excitation component amplifies measured contagion intensity by 35 to 58 percent during the COVID-19 shock and the 2022 European energy crisis relative to a transfer-entropy-only baseline; (iii) von Neumann graph entropy reaches historically extreme values 7 to 12 trading days before the peak drawdown in a Global Sovereign Bond Index. These results hold across rolling window lengths, significance thresholds, alternative entropy functionals and alternative Hawkes kernels. Full article

Financial markets are increasingly shaped by heterogeneous information sources, including trading behaviors, order-book dynamics, textual events, investor sentiment, and macroeconomic conditions. These signals are often asynchronous, noisy, partially missing, and associated with different degrees of reliability, which makes trustworthy financial risk early warning challenging. To address these issues, this study proposes an uncertainty-aware multimodal financial sensing network, termed UAMF-Net. The model treats price series, trading volume, order books, news texts, investor sentiment, and macroeconomic variables as financial sensing signals and integrates them through three task-oriented modules. First, the asynchronous multimodal temporal alignment module uses absolute time encoding, relative interval modeling, event-lag representation, temporal gating, and target-time-guided cross-scale attention to align cross-frequency financial signals according to their relevance to the prediction time. Second, the risk-aware multimodal soft fusion module estimates modality-level risk contribution and signal reliability by combining fuzzy risk membership, modality confidence weights, and cross-modal consistency constraints. Third, the uncertainty-aware risk early warning module adopts evidential learning to generate nonnegative class evidence, derive risk-category probabilities from Dirichlet parameters, estimate predictive uncertainty from total evidence strength, and jointly predict continuous risk intensity. Experimental results show that UAMF-Net achieves the best overall performance, with Accuracy, Precision, Recall, F1-score, Macro-F1, ROC-AUC, and PR-AUC reaching $0.882$, $0.864$, $0.849$, $0.856$, $0.839$, $0.941$, and $0.824$, respectively, while ECE and Brier score are reduced to $0.037$ and $0.096$. Under severe temporal asynchrony, UAMF-Net maintains an Accuracy of $0.849$, a Macro-F1 of $0.797$, and a PR-AUC of $0.774$. Under the missing multiple modalities setting, it achieves an Accuracy of $0.842$ and a Macro-F1 of $0.788$. The uncertainty analysis further shows that Risk Precision@90% reaches $0.889$. Validation on FNSPID, Daily News, and StockEmotions also confirms its generalization ability across public financial benchmarks. These results indicate that UAMF-Net improves financial risk early warning by jointly modeling temporal asynchrony, modality reliability, and predictive uncertainty. Full article

This study explores current applications of exergaming in healthcare, with a focus on how the Otago Exercise Program—a structured, evidence-based program designed to improve strength and balance in older adults—integrated into the FallSensing exergames, contributes to improving older adults’ functioning. It also aims to generate evidence to support the optimization of sensor-based technologies for more personalized and adaptable exercise interventions. Community-dwelling older adults (≥60 years) were recruited from facilities in Coimbra, Portugal, and allocated into an exergames group (IG; n = 27) and a control group (CG; n = 34). The CG maintained usual daily activities, while the IG completed an 8-week (16-session) exergame-based program. After completing the program, the CG showed a decline in functional ability, whereas the IG demonstrated significant improvements in the Step Test (p = 0.001), Four-Stage Balance Modified Test (p = 0.001), Self-Efficacy for Exercise Scale (p = 0.009), and Activities and Participation Profile Related to Mobility questionnaire (p < 0.001). Exergaming was safe and effective in enhancing functional ability, participation, and self-efficacy in older adults. However, careful consideration of exercise frequency, intensity, and participants’ age is recommended when prescribing exergame-based interventions. These results also highlight another interesting topic among physiotherapists who prescribe and monitor exergames, that technology developers should consider exercise-time monitoring systems that integrate physical (e.g., eye, facial, and mouth features) and physiological signals to enhance fatigue detection accuracy. Full article

Fall-from-height (FFH) detection is a critical component in wearable safety systems, particularly in environments where high-intensity movements can lead to frequent false positives. Conventional approaches based on simple thresholding of acceleration signals often fail to reliably distinguish FFH events from non-fall activities due to overlapping signal characteristics. This paper proposes a data-driven FFH detection method that integrates multiple complementary features into a unified score-based model. The proposed approach first performs structured peak detection to extract candidate impact events while significantly reducing the number of samples requiring further processing. Each candidate is then evaluated using pre-peak structure, post-impact stability, and pressure variation, which respectively capture structural, temporal, and physical characteristics of FFH events. Based on statistical analysis, feature-wise score contributions are designed to reflect their discriminative strength, and the final FFH decision is performed using an additive scoring mechanism. This formulation enables flexible handling of ambiguous cases while preserving strong FFH characteristics. Experimental results demonstrate that the proposed method maintains 100% recall at the selected decision threshold while significantly reducing false positives from non-FFH activities. In addition, the peak detection stage reduces more than 99% of raw samples, enabling efficient on-device processing suitable for wearable systems. The proposed method also includes quantitative analysis of latency characteristics. Although FFH inference latency is influenced by asynchronous pressure sensing, the delay remains bounded and predictable, and most detections are completed within a practical time range for real-time wearable safety applications. Overall, the proposed method achieves a practical balance between detection sensitivity, false-positive suppression, computational efficiency, and real-time feasibility, demonstrating its applicability to wearable safety systems. Full article

Real-time, non-contact velocity measurement of hot-rolled bars is critical for metallurgical process control, but conventional laser Doppler velocimetry (LDV) systems often fail in these environments. The intense broadband thermal radiation from targets up to 1000 °C, coupled with severe surface depolarization, overwhelms weak scattered signals in high-speed (up to 40 m/s) rolling zones. To address this issue, we developed a fully integrated, thermal-radiation-resistant LDV sensing system. Hardware optimization was achieved by eliminating polarized-light transmission and adopting a parallel-beam design, which significantly enlarges the laser overlap area and increases detection depth. Furthermore, a 1550 nm laser (100 mW) was coaxially combined with a 10 nm narrow-band filter to isolate the thermal background and boost signal strength. A customized workflow utilizing continuous Fourier transform (CFT) spectral refinement and energy centroid estimation was implemented to precisely extract the true Doppler shift. Performance evaluations show the system achieves an excellent signal-to-noise ratio (SNR) of 29,532. Allan variance analysis confirms a stable detection sensitivity of 0.003 m/s (0.1 s integration time), a local short-to-medium-term optimal limit of 1.6 × 10⁻⁴ m/s, and a statistical accuracy of 0.005 m/s. Finally, the system was successfully deployed on an industrial rolling mill production line. It provided reliable velocity feedback for mill speed adjustment, achieving a near-zero-tension rolling process and fundamentally resolving workpiece dragging, squeezing, and steel pile-up. Full article

Hybrid RF-FSO-THz communication systems are promising candidates for future Internet of Things (IoT) and 6G networks because they combine the robustness of radio frequency links, the high-capacity potential of Free-Space Optical communications, and the ultra-wideband capabilities of terahertz transmission. Adaptive channel selection in such systems depends on multiple correlated environmental and physical-layer variables, including distance, rain intensity, humidity, visibility, turbulence strength, signal-to-noise ratio, channel capacity, and energy-efficiency metrics. This paper presents a physics-supported benchmark framework for supervised adaptive channel selection in hybrid RF-FSO-THz systems and systematically investigates the impact of linear and nonlinear dimensionality-reduction techniques on predictive performance, statistical robustness, computational complexity, and physical interpretability. A multi-scenario dataset comprising 5000 samples was generated using calibrated RF, FSO, and THz propagation models under clear, rain, fog, and worst-case environmental conditions. Principal Component Analysis (PCA) and Kernel PCA were evaluated together with Random Forest, Support Vector Machines (SVMs), XGBoost, Gradient Boosting (GB), Multi-Layer Perceptron (MLP), Logistic Regression, and Decision Trees. The results demonstrate that PCA preserves nearly all predictive capabilities while reducing the original 33-dimensional feature space by approximately 81.8%, maintaining accuracies close to 97–98% with the best-performing classifiers. Statistical significance analysis confirms that PCA introduces only modest degradations, whereas Kernel PCA consistently reduces the predictive performance while increasing memory requirements and inference latency. Additional environmental-only validation experiments indicate that adaptive channel selection remains highly learnable even when only pre-selection environmental descriptors are available, partially mitigating concerns regarding self-consistency bias. Overall, the results suggest that PCA provides an advantageous compromise among predictive accuracy, computational efficiency, statistical robustness, and physical interpretability for supervised adaptive channel selection in physics-supported hybrid wireless communication systems. Full article

This paper considers a multi-period two-sided asymmetric information model with infinitely long-lived sellers and short-lived buyers. I assume that two exogenously given qualities are offered in the market. Each period, a consumer, who is uncertain about the quality of the offered product, observes her pairwise matched seller’s price and a noisy signal of quality that cannot be manipulated by the seller. Prices are fixed and it is common knowledge that consumers are not willing to pay a high price for the low-quality product. A matched seller with a low-quality good can choose to be either honest (by charging the lower market price) or dishonest (by charging the higher price). Sellers’ incentives to misrepresent quality depend on how current trade outcomes affect future access to consumer traffic. I show that the strength of the informational role of prices is non-decreasing in the intensity of competition for future consumer traffic in equilibrium and that consumers do not benefit from more intense competition. Full article

Accurate assessment of vitreous inflammation is essential for the diagnosis, monitoring and management of uveitis. Traditionally, vitritis has been evaluated using subjective clinical grading systems based on vitreous haze and cellular infiltration, which are limited by interobserver variability and poor reproducibility, particularly in cases of mild or subclinical inflammation. In recent years, advances in ocular imaging have enabled the development of more objective, quantitative approaches. Ultra-widefield imaging, optical coherence tomography (OCT) and ultrasound-based techniques have provided new insights into structural alterations within the vitreous. In parallel, automated image analysis and artificial intelligence (AI)-based methods have improved the detection and quantification of inflammatory biomarkers, including vitreous hyperreflective foci and signal intensity-based metrics. Despite these advances, important limitations remain, including a restricted field of view, a lack of standardized segmentation algorithms and an incomplete representation of the entire vitreous cavity. No single modality currently provides a comprehensive and fully reproducible assessment of vitreous inflammation. This review summarizes current qualitative and quantitative methods for evaluating vitreous inflammation, highlighting their respective strengths and limitations. In addition, emerging diagnostic strategies, including multimodal imaging integration, AI-driven analysis and molecular biomarker profiling, are discussed as potential tools to improve accuracy, standardization and clinical applicability. The transition from subjective grading toward objective quantification of inflammatory burden represents a key step in advancing both clinical management and research in ocular inflammatory diseases. Full article

Background: Although early mobilization has been shown to improve clinical outcomes after intensive care unit (ICU)-acquired weakness, its implementation remains limited in routine clinical practice. This study aimed to evaluate the feasibility, safety, and preliminary clinical outcomes of a mobile application-based rehabilitation program in non-ambulatory patients during the early ward phase following ICU discharge. Methods: This prospective single-arm pilot trial included adult patients (≥19 years) who had received ICU care and demonstrated limited ambulatory function, defined as Functional Ambulatory Category (FAC) ≤3. Participants received an individualized, application-guided exercise program comprising two daily sessions over two weeks. Primary outcomes were programmatic feasibility, safety, and patient satisfaction. Rehabilitation compliance was quantified using application usage logs and categorized as high (≥50%) or low (<50%). Secondary functional outcomes, such as Medical Research Council Sum Score (MRC-SS), ICU Mobility Scale, FAC, muscle strength measures, health-related quality of life, and pain scores, were assessed at baseline, week 1, and week 2. Results: Of the 25 initially enrolled patients, 5 dropped out due to clinical status changes or transfers, yielding a retention rate of 80.0%. For the 20 analyzed patients (mean age 52.7 ± 13.9 years; 45% male), the overall mean rehabilitation compliance was 40.6%. No serious adverse events related to the intervention were reported, and overall patient satisfaction and application usability were high. Progressive increases in exercise intensity and training levels were observed throughout the intervention period. Significant improvements over time were found in MRC-SS, ICU Mobility Scale, FAC, grip strength, health-related quality of life, and pain scores (all p < 0.05). Although compliance-based recovery trajectories were confounded by small subgroup sizes and baseline clinical imbalance, exploratory analyses nonetheless identified statistically significant time × compliance interaction effects for MRC-SS and straight leg raise performance. Conclusions: This pilot study demonstrates that a mobile application-based rehabilitation program is a feasible and safe approach to implement in deconditioned patients after ICU discharge. These preliminary functional recovery trajectories provide encouraging signals, suggesting that this digital platform may serve as a potential adjunct to conventional care. Rigorous, randomized controlled trials are required to confirm its definitive clinical efficacy and scalability. Full article

Producing high-conductivity aluminum conductors for power transmission involves 23 trace elements and multiple interconnected thermo-mechanical stages. The ultra-low alloying levels required to preserve high electrical conductivity create a narrow compositional window and highly imbalanced distributions, which hinder traditional data-driven learning. Here, we developed a physics-guided machine-learning framework based on 4458 valid industrial production records to predict tensile strength and electrical resistivity. In addition to raw composition and process parameters, we introduce ratio descriptors (e.g., Fe/Si and Al/Si) and propose a physics-informed metric termed the Equivalent Solute–Heat Index (ESHI) to couple key solute chemistry (Si, Fe, B) with normalized thermal-history intensity. Fe and Si primarily influence resistivity through impurity/solute scattering, while B mainly affects microstructural uniformity via grain refinement. Incorporating ESHI as an augmented signal into the best-performing XGB surrogate markedly improves generalizability, increasing the tensile strength R² from 0.75 to ~0.92. SHAP analysis reveals that ESHI dominates the decision logic by modulating both targets with metallurgically interpretable mechanisms: solute-controlled scattering and thermal history-traced second-phase evolution that stabilizes the microstructure. NSGA-III was further employed to map the Pareto front and identify composition–process combinations that optimize the strength–conductivity trade-off, enabling improved mechanical reliability while minimizing resistive losses in practical power-transmission applications. Experimental validation on industrial wires confirms this reliability. Full article

Indoor localization in GNSS-denied environments remains a significant challenge due to the low sampling frequency and high variability of wireless signal measurements. This paper presents a wireless communication-based indoor localization method that integrates Wi-Fi received signal strength indication (RSSI) measurements with optical initialization and inertial sensor fusion. The proposed approach eliminates the need for labor-intensive fingerprinting and specialized infrastructure by leveraging existing Wi-Fi networks. Optical pose estimation using ArUco markers provides accurate initial position and orientation, enabling alignment between sensor coordinate systems and reducing inertial drift. During tracking, inertial measurements compensate for motion between sparse Wi-Fi observations by virtually translating historical RSSI samples, allowing statistically consistent averaging and improved distance estimation. A simplified factor graph framework is employed to fuse heterogeneous measurements while maintaining computational efficiency suitable for real-time operation on mobile devices. Experimental validation using a robot-based ground-truth reference system demonstrates sub-meter localization accuracy with an average positioning error of approximately 0.40 m. The proposed method provides a low-cost and scalable solution for indoor positioning and navigation applications such as access-controlled environments, exhibitions, and large public venues. Full article

Background/Objectives: To evaluate the influence of different DIXON contrasts on the quality of subtraction images in dynamic breast MRI using maximum intensity projections (MIPs). Methods: This retrospective study included n = 40 women (median age: 53.5 years, range 23–83) undergoing clinically indicated breast MRI (3T). For each MRI examination, two independent readers individually evaluated GBCA-enhanced subtraction MIPS for different timepoints (n = 5) and DIXON contrasts (n = 4) per breast, resulting in a total of 800 individual evaluations. Evaluations comprised (a) qualitative measures, using Likert-scores for artefact strength, breast parenchyma visibility, lesion visibility and reading confidence; and (b) signal intensity, measured in three regions of interest with the apparent signal-to-noise ratio (aSNR) and apparent contrast-to-noise ratio (aCNR) calculated. The evaluation results were analysed to identify differences between DIXON contrasts. Results: The “only water” DIXON contrast at ~120s after GBCA injection achieved the highest lesion conspicuity and reading confidence scores and lowest artefact scores; however, its performance was not statistically significant (p > 0.05) compared to the “in-phase” and “opposed-phase” subtractions. The aCNR at the second timepoint was slightly, but not significantly (p > 0.05), lower than the first timepoint, whilst aSNR increased significantly from the first to second timepoint in all contrasts. Conclusions: Subtraction MIPs derived from the “only water” DIXON contrast achieved the highest qualitative scoring for lesion conspicuity and confidence, with the aSNR increasing and aCNR decreasing between the first and second timepoints. Full article

A dual-mode electrical–optical nanocomposite hydrogel is developed by integrating carboxyl-modified upconversion nanoparticles (UCNPs-COOH) and quaternized chitosan (CQAS) into a polyacrylamide (PAAm) covalent network. The hydrogel exhibits high optical transparency (>90% in the visible region), excellent mechanical properties (fracture strain of 1742%, tensile strength of 0.85 MPa, toughness of 6.57 MJ/m³), and robust adhesion to various substrates. The synergistic covalent–noncovalent hybrid network enables efficient energy dissipation, while CQAS-enhanced dispersion of UCNPs significantly improves upconversion luminescence intensity and stability, as evidenced by prolonged fluorescence lifetime from 0.564 ms to 0.691 ms at 539 nm. Leveraging distinct electrical and optical signal transduction pathways, the hydrogel functions as a highly sensitive resistive strain sensor with multistage gauge factors up to 13.85 and excellent cyclic stability over 1200 loading–unloading cycles at 100% strain for human motion monitoring. It also serves as a ratiometric optical pH sensor over a broad range (pH 1–13) based on phenolphthalein-sensitized upconversion luminescence, with excellent repeatability. By integrating real-time resistance responses with optical readouts within a single soft material, this work demonstrates a reliable dual-mode sensing strategy for simultaneous mechanical and chemical monitoring, holding promise for wearable electronics, smart healthcare, and environment-responsive sensing systems. Full article