Search Results (1,219)

A novel facial landmark localization method, termed TSDM, is proposed by integrating T-shaped features with the Supervised Descent Method (SDM). Facial landmark localization is critical for driver fatigue and attention detection in intelligent cockpits. Traditional methods lack accuracy and robustness in complex in-cabin environments such as varying illumination and head pose changes, while deep learning approaches are computationally expensive on resource-constrained vehicle platforms. The T-shaped feature well matches facial geometry and enhances feature representation. T-shaped features are selected via AdaBoost for robust face detection, and SDM is then used to locate 68 facial landmarks. Experiments show that TSDM achieves higher accuracy, lower false-positive rates, and better efficiency than traditional methods, including Haar and LBPH. It also exhibits stronger robustness and better real-time performance than several lightweight deep learning models (such as 3D-aware methods and SAN) on CPU-only platforms, while achieving comparable or higher localization accuracy. Experimental results show that TSDM achieves a face detection rate of 97.43% and a normalized mean error (NME) of 3.4% on standard datasets. The proposed method provides a practical solution for driver state monitoring in resource-limited vehicular environments. Full article

EEG-based driver fatigue detection faces three signal-level challenges: inter-subject spectral variability, coupled frequency–spatial–temporal dynamics that existing methods process independently, and dependence on a single labeling scheme. This paper presents DCAMNet, a lightweight CNN (12.3 K parameters) that addresses these challenges through three end-to-end blocks. An SE-driven dynamic convolution block adapts spectral sensitivity per sample via input-dependent kernel weighting—applied here for the first time to fatigue detection. A spatial convolution block encodes electrode-level cortical patterns, and a temporal attention block captures fatigue dynamics through windowed variance descriptors with group-wise attention scoring. DCAMNet was evaluated on SEED-VIG (PERCLOS labels) and MESD (reaction-time labels) under both subject-mixed and leave-one-subject-out (LOSO) protocols. Under LOSO cross-validation—the operationally relevant test that eliminates within-subject information leakage and simulates deployment on unseen drivers—DCAMNet achieved 85.43% accuracy on SEED-VIG with a 2.86-point advantage over the strongest baseline, and $79\pm 5$% accuracy on MESD with a 3-point advantage. As upper-bound estimates under the subject-mixed protocol, accuracy reached 97.47% (SEED-VIG) and 96.52% (MESD). With 1.35 ms inference latency on a standard GPU, the compact architecture suggests potential suitability for real-time embedded deployment, although on-device validation on representative automotive hardware remains necessary. Full article

As critical equipment in the petroleum industry, coiled tubing is prone to safety hazards, including stress concentrations and fatigue failure, under complex operating conditions. An online enhanced metal magnetic memory detection method was employed to reduce noise in surface magnetic field signals from tubing subjected to 35 MPa of internal pressure across different fatigue cycles. Conventional signal processing methods have difficulty effectively extracting characteristic magnetic field signals in high-noise environments; therefore, a comprehensive comparison of the noise reduction effectiveness of five common signal processing techniques in stress-distorted regions was conducted, an in-depth analysis of the limitations of different methods was performed, and a hybrid noise reduction framework combining wavelet threshold denoising (WTD) and sequential variational modal decomposition (SVMD) was established. Concurrently, the whale optimization algorithm (WOA), which possesses global search capabilities and demonstrates good adaptability to multi-parameter coupling issues in hybrid denoising frameworks, was innovatively proposed for key parameter optimization. Using fuzzy entropy (FE) as an evaluation metric, the experimental results demonstrated that magnetic field signals in all directions achieved at least a 1.03% reduction in FE and a minimum increase of 33.1% in integrated side lobe ratio (ISLR). This provided effective technical support for reliably detecting stress-distortion zones on coiled-tubing surfaces and established the engineering necessity of implementing preventive maintenance. Full article

Gingival recession is commonly linked to alveolar bone dehiscence, inflammatory burden, traumatic brushing, or excessive orthodontic forces. However, recession is also observed in some patients despite apparently mild or “biologically acceptable” loading, particularly in thin periodontal phenotypes. Here, we propose the Gingival Creep Failure Theory, a hypothesis-driven conceptual framework in which gingival soft tissues undergo time-dependent viscoelastic deformation (creep) under sustained or repetitive tensile microstrain. Over time, accumulated deformation and microstructural fatigue may reduce recoil capacity and shift the gingival margin apically once tissue-level tolerance is exceeded. Gingival connective tissue is modeled as a fiber-reinforced, fluid-rich viscoelastic composite whose response depends on collagen architecture, cross-linking, proteoglycan-mediated hydration, and vascular support. In thin phenotypes characterized by reduced connective tissue volume and altered extracellular matrix (ECM) organization, creep progression is hypothesized to accelerate, lowering the threshold at which fatigue-related microdamage translates into clinically detectable marginal migration. Evidence from collagenous connective tissue biomechanics supports the plausibility that sub-failure sustained or cyclic loading can produce cumulative deformation and incomplete recovery; however, direct creep–fatigue data for human gingiva remain limited, underscoring the need for targeted validation studies. This hypothesis integrates soft tissue mechanics with periodontal phenotype biology and orthodontic loading patterns and proposes creep and microstructural fatigue as plausible time-dependent contributors to gingival recession in susceptible phenotypes. Because direct in vivo gingival strain and creep–fatigue measurements remain limited, the model should be interpreted as hypothesis-generating and in need of targeted clinical and experimental validation. Full article

The Officer on Watch (OOW) is critical to ensuring the safety of the vessel, cargo, and crew during navigation. To reduce maritime accidents caused by unsafe behaviors or fatigue, this paper proposes a dual-branch detection system based on Spatial–Temporal Graph Convolutional Networks (ST-GCN): BODY-ST-GCN for pose-based behavior recognition and FACE-ST-GCN for facial state analysis. For spatial modeling, a Triple Graph Fusion (TGF) strategy is introduced to integrate static, adaptive, and attention graphs, enhancing the representation of skeletal and facial keypoints. For temporal modeling, BODY-ST-GCN incorporates a Three-Scale Parallel Temporal Convolutional Network (TSP-TCN) to capture multi-scale motion dynamics, while FACE-ST-GCN uses a Temporal Adaptive Module (TAM) to extract stable facial state features. Furthermore, a joint risk classification mechanism categorizes OOW duty states into four hierarchical levels: Safe, Early Fatigue Warning, High Fatigue Risk, and Emergency. This mechanism enables continuous, real-time monitoring and dynamic assessment. Experiments demonstrate that BODY-ST-GCN and FACE-ST-GCN achieve macro average precisions of 0.969 and 0.947, respectively, outperforming the baseline ST-GCN by 6.4% and 14.9%, providing reliable technical support for onboard safety management. Full article

Distracted and fatigue-related driving behaviors are major causes of road traffic accidents, creating an urgent need for reliable driver monitoring systems. Vision-based detection methods have garnered widespread attention due to their low cost of deployment and practical applicability. However, existing lightweight models often suffer from limited global contextual perception and insufficient preservation of fine details. Motivated by these challenges, this study introduces an improved distracted and fatigue-related driving behavior detection model, YOLOv12-LAD, built on the YOLOv12 architecture. The proposed framework integrates a Large Separable Kernel Attention module (LSKA) to enhance global contextual perception, an Adaptive Downsampling module (ADown) to mitigate information loss during feature compression, and a Dynamic Sampling module (DySample) to enable content-adaptive feature reconstruction and improve multi-scale behavior representation. Experimental results show that YOLOv12-LAD achieved 97.5% precision, 96.3% recall, and 98.4% mAP@50 with only 2.5 million parameters, 6.2 GFLOPs, and an inference speed of 249 FPS. Ablation studies, comparisons with representative models, cross-dataset evaluation, and real-vehicle tests further verify the effectiveness and robustness of the proposed method. The proposed method demonstrates strong performance while maintaining computational efficiency, making it suitable for real-time vision-based driver monitoring applications. Full article

Owing to the intricate mechanical coupling characteristics and the considerable difficulty in extracting synergistic spatio-temporal features from high-dimensional sensor data under fluctuating alternating loads, this study proposes a robust anomaly detection framework that combines Normalized Mutual Information (NMI) and Spatio-Temporal Graph Neural Networks (STGNN). First, NMI is utilized to quantify the nonlinear physical coupling intensity among multi-source sensors, thereby filtering out weakly correlated noise and reconstructing the spatial topological structure of the coupler system. Subsequently, a deep learning architecture incorporating Graph Convolutional Networks (GCN), Gated Recurrent Units (GRU), and one-dimensional convolutional residual connections is developed to capture the dynamic evolutionary characteristics of equipment states across both spatial interactions and temporal sequences. Finally, based on the model’s health-state predictions, a moving average algorithm is introduced to smooth the residual sequences, and an anomaly early-warning baseline is established in conjunction with the 3σ criterion. Experimental validation conducted using field service data from heavy-haul trains demonstrates that, compared to conventional serial CNN and Long Short-Term Memory (LSTM) models, the proposed method exhibits superior fitting performance and robustness against noise, effectively reducing the false alarm rate within normal working intervals. In a real-world case study, the method successfully identified variations in spatial linkage features induced by local damage and triggered timely alerts. Notably, the spatial alarm nodes were highly consistent with the fatigue crack initiation sites identified through on-site magnetic particle inspection. This study provides a viable data-driven analytical framework for the condition monitoring and anomaly identification of critical load-bearing components in heavy-haul trains. Full article

Background: Percussive massage therapy (PMT) with handheld massage guns is widely used to support recovery and flexibility, but the short-term behavior of skeletal muscle contractile properties and the relative contribution of application duration versus frequency remain unclear. This study investigated the 10 min post-intervention time course of tensiomyography (TMG)-derived contractile properties of non-fatigued vastus medialis (VM) after clinically realistic PMT protocols and examined whether longer duration is associated with persistent deviations from baseline than frequency. Methods: In a two-session, within-subject repeated-measure design, 32 participants completed four PMT conditions to the VM (35 Hz–3 min, 35 Hz–6 min, 45 Hz–3 min, and 45 Hz–6 min). TMG parameters (Td, Tc, Ts, Tr, and Dm) were recorded at baseline and repeatedly over 10 min post-intervention. Linear mixed-effect models with frequency and duration as fixed factors and time as continuous and categorical were used to characterize temporal patterns, with emphasis on effect sizes and consistency across parameters. The fixed protocol order (35 Hz in session one, 45 Hz in session two, 3 vs. 6 min assigned to contralateral legs) means that frequency was confounded with session and duration with leg side. Results: Compared with the 3 min protocols, the 6 min protocols were associated with slightly higher Td and Ts, a modest increase in Tr and a slightly greater Dm (e.g., Dm + 0.55 mm), whereas Tc showed no clear duration effect. Across conditions, Td increased immediately after PMT, Tc remained elevated for most of the first 8 min, Ts increased from mid to late post-intervention, Tr changed inconsistently, and Dm was reduced relative to baseline for most of the 10 min period. Differences between 35 and 45 Hz were small and non-significant for all TMG parameters. Conclusions: Clinically realistic PMT protocols at 35–45 Hz in non-fatigued muscle induce small but statistically detectable, duration-sensitive changes in TMG-derived contractile behavior over approximately 10 min. Within the constraints of the fixed, non-randomized design and the small effect sizes observed, these findings support viewing massage gun use as a recovery-oriented adjunct that subtly modulates contractile dynamics, rather than as a strong, standalone performance-enhancing stimulus. Full article

Laser-based Powder Bed Fusion (LPBF) enables the fabrication of complex metal components but often results in high porosity and microdefect densities, compromising fatigue performance despite acceptable static properties. Standard fatigue characterisation methods are time-consuming and costly and yield scattered results due to defect-induced brittleness and residual stresses. This study investigates the application of thermographic techniques as a rapid alternative for evaluating the intrinsic fatigue behaviour of tensile coupons fabricated by LPBF employing AISI 316L steel. By monitoring surface temperature during stepwise static monotone and fatigue loading, thermographic methods aim to detect early hints of heat dissipation associated with microdamage initiation. Approaches based on temperature harmonic analysis have been implemented, allowing near-real-time and full-field mapping of stress distribution and damage development. Results show that harmonic metrics correlate with the material state and effectively track the thermoelastic effect-induced temperature changes. Some evidence is found regarding the onset of intrinsic heat dissipation, which needs to be confirmed by more focused and extensive experimental tests. Full article

Background/Objectives: Radiation therapy is a common treatment modality for non-small-cell and small-cell lung cancer that can be associated with considerable side effects, mainly reactions of healthy tissues in the radiation field. Radiation therapy may lead to significant fatigue, which can potentially be mitigated by maintaining or increasing physical activity during treatment. Since achieving this goal may be a challenge for patients, they may benefit from a mobile application reminding them daily to perform a predefined number of steps. Such a reminder app will be investigated prospectively in a phase 2 trial. The current APART-LUNG study (NCT07380815) is a mandatory study for designing the prospective trial. Methods: The main objective of the APART-LUNG (exploratory non-interventional) study is to report patterns of physical activity during radiation therapy for lung cancer patients and generate hypotheses based on our findings. Our primary endpoint is the within-patient difference in weekly average steps per wear hour of the smartphone (week 5 minus week 1 of radiation therapy), and our secondary aim is to estimate differences in operational measures (wear time of the smartphone) between week 5 and week 1. The sample size of approximately 20 patients (full analysis set) allows us to detect a moderate-to-large standardized within-patient difference and is driven by feasibility and the intent to obtain preliminary estimates of effect size and variability. The results of the APART-LUNG study will be very important for appropriately designing a phase 2 trial. Full article

To address the growing demand for sustainable road infrastructure development and resolve technical bottlenecks in reclaimed asphalt pavement (RAP) recycling, this study optimized the performance of recycled asphalt mixtures (RAMs) and validated their engineering applicability for field construction. RAM specimens were prepared using 5-year and 10-year aged RAP from Ningxia, with a constant RAP content of 30%. Laboratory tests including high-temperature rutting, moisture susceptibility, low-temperature cracking, dynamic modulus, and four-point bending fatigue were performed to determine the optimal mix proportion. Fourier Transform Infrared Spectroscopy (FTIR) and Thin-Layer Chromatography-Flame Ionization Detection (TLC-FID) were employed to reveal the regeneration mechanism of waste engine oil (WEO). Results showed that WEO modified the functional groups and four fractions of asphalt, optimizing its colloidal structure, while excessive WEO compromised high-temperature stability. The optimal WEO contents were 4% for RAP (5Y) and 8% for RAP (10Y), which significantly enhanced the overall performance of RAM to adapt to Ningxia’s climate. This study provides technical support for sustainable road infrastructure in arid and semi-arid regions. Full article

Background: Saudi Arabia has made significant progress toward malaria elimination; however, imported cases continue to occur, particularly in the southwestern regions. This study aimed to describe the clinical characteristics and outcomes of patients with malaria in the Aseer Region, Saudi Arabia. Methods: A retrospective observational study was conducted at Khamis Mushait General Hospital, Aseer Region, Saudi Arabia, including all patients with malaria from January 2022 to December 2025. Demographic, clinical, laboratory, and outcome data were extracted from the electronic medical records. Severe malaria was defined according to the World Health Organization criteria. Multivariate logistic regression using Firth’s penalized maximum likelihood estimation was performed to identify independent predictors of severe malaria (≥1 WHO criterion). Statistical analysis was performed using R software (version 4.2.1). Results: A total of 311 patients were included, predominantly male (90.0%), with a mean age of 28.8 ± 11.3 years. Ethiopian nationals comprised nearly half the cases (48.2%), followed by Saudi (16.4%) and Yemeni (15.1%) nationals. Plasmodium vivax was the most common species (51.1%), followed by Plasmodium. falciparum (40.2%). Fever was the most frequent symptom (89.4%), followed by fatigue (50.8%), chills (46.9%), and vomiting (39.5%). Low parasitemia (<1%) was the most frequent finding (33.8%), followed by moderate (27.3%) and mild (18.3%) levels, while high (4.2%) and very high parasitemia (1.9%) were uncommon. Severe malaria (≥1 criterion) was diagnosed at 43.7%, with severe anemia (26.0%) and jaundice (23.2%) being the most frequent WHO severity criteria. Notably, 84% of the cases occurred during 2024–2025, indicating a recent outbreak, with a sharp peak of 43 cases in October 2024. Multivariate logistic regression identified two independent predictors of having at least one WHO severity criterion: higher parasitemia level (adjusted OR = 1.70 per 1% increase, 95% CI: 1.40–2.11, p < 0.001) and non-Saudi nationality (adjusted OR = 2.40, 95% CI: 1.10–5.62, p = 0.027). Conclusions: Malaria in the Aseer Region predominantly affects young adult male expatriates, suggesting its imported nature. The predominance of P. vivax represents a shift from historical patterns. Parasitemia level and being of non-Saudi nationality independently predict severe malaria and may therefore support risk stratification and clinical decision-making. The dramatic case surge in 2024–2025 highlights regional vulnerability to outbreaks despite control progress. These findings support enhanced screening for at-risk populations, maintenance of clinical capacity for severe malaria management, and robust surveillance systems for early outbreak detection. Full article

Heart rate variability (HRV) has emerged as a key biomarker for assessing autonomic nervous system activity, stress, fatigue, and emotional states. With the rapid development of wearable sensor technologies, HRV analysis has expanded from clinical environments to real-world, continuous monitoring. This review summarizes current applications of HRV metrics in wearable devices, including fitness tracking, mental stress assessment, sleep quality evaluation, and early detection of physiological or psychological disorders. Recent advances in photoplethysmography (PPG)-based HRV estimation have enabled noninvasive and user-friendly measurement, though challenges remain in accuracy under motion and variable environmental conditions. We also discuss methodological considerations, such as artifact correction, data segmentation, and the integration of HRV with other biosignals for multimodal analysis. Emerging research suggests that combining HRV with metrics such as respiration rate, skin conductance, and accelerometry can enhance robustness and interpretability in dynamic settings. Finally, future directions are proposed toward personalized health analytics, emotion-aware computing, and real-time adaptive feedback systems. This review highlights the growing potential of wearable HRV analysis as a foundation for preventive healthcare and human–machine symbiosis. Full article

Toxoplasmosis is a widely distributed zoonosis caused by the protozoan parasite Toxoplasma gondii. Infection during pregnancy is a major public health concern due to its potential impact on both maternal health and fetal development. Early detection of maternal infection is critical to prevent adverse outcomes; however, maternal signs are often subtle, non-specific or absent, complicating timely diagnosis. This scoping review aimed to map and synthesise existing evidence on early maternal signs, pregnancy and foetal outcomes, frequently assessed risk factors, and diagnostic approaches of toxoplasmosis in expectant mothers in Africa. The review was done in accordance with the PRISMA-ScR guidelines. A literature search of PubMed, Scopus, ResearchGate, and Google Scholar was performed to identify studies published between 2000 and 2025. Retrieved records were managed using Zotero (version 8.0.4) for deduplication and screening. Only English-language studies conducted in Africa and reporting relevant maternal or clinical data were included. A total of 28 cross-sectional studies were included. Lymphadenopathy (25.0%) was the most frequently reported maternal early sign, followed by flu-like illness, asymptomatic infection, low-grade or mild fever, and fatigue or malaise (each 10.7%). Congenital anomalies (50.0%) and miscarriage or spontaneous abortion (42.9%) were the most commonly reported foetal and pregnancy outcomes. Frequently reported risk factors were exposure to cat faeces (57.1%) and ingestion of undercooked or raw meat (42.9%). Diagnostic approaches were commonly enzyme-based immunoassays (78.6%), with limited use of RDTs and molecular methods. These findings suggest the need for improved early detection and prevention strategies in high-risk, low-resource African settings. Enhancing routine screening, health education, and access to appropriate diagnostics are considered. Future studies should consider adopting standardised reporting and integrating sensitive, affordable, rapid diagnostic approaches to enhance early detection and reduce the burden of congenital toxoplasmosis. Full article

Wind turbine blades are subjected to complex environmental conditions during long-term operation, which may lead to structural degradation and performance loss. To ensure structural integrity, fatigue testing prior to deployment is essential. This paper proposes a vision-based method for measuring the full-cycle breathing deformation of wind turbine blades during fatigue testing. The method captures dynamic image sequences of the blade’s hotspot cross-section using industrial cameras and employs a feature-based template matching approach to reconstruct the three-dimensional coordinates of target points. Through coordinate transformation, the deformation trajectories are obtained, enabling quantitative analysis of the blade’s dynamic responses in both flapwise and edgewise directions. A dedicated hardware–software system was developed and validated through full-scale fatigue experiments. Quantitative comparison with strain gage measurements shows that the proposed method achieves mean absolute deviations of 0.84 mm and 0.93 mm in two independent experiments, respectively, with closely matched deformation trends under typical loading conditions. These results demonstrate that the proposed method can reliably capture the global deformation behavior of the blade with millimeter-level accuracy, while significantly reducing instrumentation complexity compared to conventional contact-based approaches. The proposed method provides an effective and practical solution for full-field dynamic deformation measurement in blade fatigue testing, offering strong potential for structural health monitoring and early damage detection in wind turbine systems. Full article