Search Results (219)

Environmental changes and sensor aging can cause sensor drift in sensor array responses (i.e., a shift in the measured signal/feature distribution over time), which in turn degrades gas classification performance in real-world deployments of electronic-nose systems. Previous studies using the UCI Gas Sensor Array Drift Dataset as a benchmark reported promising drift compensation results but often lacked robust statistical validation and may overcompensate for drift by suppressing class-discriminative variance. To address these limitations and rigorously evaluate improvements in sensor-drift compensation, we designed two domain adaptation tasks based on the UCI electronic-nose dataset: (1) using the first batch to predict remaining batches, simulating a controlled laboratory setting, and (2) using Batches 1 through $n-1$ to predict Batch n, simulating continuous training data updates for online training. Then, we systematically tested three methods—our semi-supervised knowledge distillation method (KD) for sensor-drift compensation; a previously benchmarked method, Domain-Regularized Component Analysis (DRCA); and a hybrid method, KD–DRCA—across 30 random test-set partitions on the UCI dataset. We showed that semi-supervised KD consistently outperformed both DRCA and KD–DRCA, achieving up to 18% and 15% relative improvements in accuracy and F1-score, respectively, over the baseline, proving KD’s superior effectiveness in electronic-nose drift compensation. This work provides a rigorous statistical validation of KD for electronic-nose drift compensation under long-term temporal drift, with repeated randomized evaluation and significance testing, and demonstrates consistent improvements over DRCA on the UCI drift benchmark. Full article

This research examines the underlying processes of public stigma toward visible facial differences (VFDs) by focusing on gaze behavior. Past research showed that a VFD influences the visual processing of faces, leading to increased attention to the VFD area at the expense of internal features (i.e., eyes, nose, mouth). Since these features primarily convey affective information, this pre-registered study investigates whether this bias also affects pain perception. In an eye-tracking task, participants (N = 44) viewed faces that either did or did not display a VFD located in a peripheral area of the face, and that either did or did not express pain, while their gaze behavior was being recorded. Participants then rated perceived pain intensity for each face. Results showed that VFDs diverted attention toward peripheral features and away from internal, pain-relevant features of the face. Surprisingly, participants rated faces with VFDs as experiencing more pain, regardless of whether pain was actually expressed. This suggests that, despite gazing less at facial expressions, observers inferred pain based on task-irrelevant features, likely stereotypes related to the VFD. These findings provide insights into how people with VFDs are perceived and how their emotions are interpreted. Full article

Detecting aircraft in Airport Surface Movement Radar (SMR) imagery presents a unique challenge rooted in the conflict between object symmetry and data asymmetry. While aircraft possess strong structural symmetry, their radar signatures are often sparse, incomplete, and highly asymmetric, leading to target loss and position jitter in traditional detection algorithms. To overcome this, we introduce SWCR-YOLO, a keypoint detection framework designed to learn and enforce the target’s implicit structural symmetry from its imperfect radar representation. Our model reconstructs a stable aircraft pose by localizing four keypoints (nose, tail, wingtips) that define its symmetric axes. Based on YOLOv11n, SWCR-YOLO incorporates a MultiScaleStem module and wavelet transforms to effectively extract features from the sparse, asymmetric scatter points, while a Multi-Scale Convolutional Attention (MSCA) module refines salient information. Crucially, training is guided by a Geometric Regularized Keypoint Loss (GRKLoss), which introduces a symmetry-based prior by imposing angular constraints on the keypoints to ensure physically plausible pose estimations. Our symmetry-aware approach, on a real-world SMR dataset, achieves an mAP50 of 88.2% and reduces the trajectory root mean square error by 51.8% compared to MTD-CFAR pipeline methods, from 8.235 m to 3.968 m, demonstrating its effectiveness in handling asymmetric data for robust object tracking. Full article

Accurate recognition of each cattle in group environments is essential for modern precision livestock management. This study proposed a multi-object cattle recognition method based on deep learning, enabling precise recognition in feeding passages. A dataset comprising facial images from 135 cattle was constructed, and a data augmentation strategy tailored to cattle facial characteristics was designed to enhance model generalisation. The YOLOv8n network was selected from a comparative experiment and further optimised. For multi-object bounding box regression, the standard CIoU loss was replaced by the MPDIoU loss, improving the mAP50 by 5.4% through optimised corner distance computation. In addition, a coordinate attention mechanism was embedded within the C2F module to strengthen the model’s spatial perception of key facial regions such as the eyes and nose, resulting in a 5.8% improvement in recognition precision. A comparative experiment between image-level segmentation and cattle-level segmentation datasets was carried out, and the proposed method was further validated on an untrained external test set collected from actual feeding Passages. The results demonstrate that, even under challenging conditions such as occlusion and illumination variation, the improved model achieved a classification accuracy of 88% while maintaining an average inference speed of 96.9 frames per second. This non-invasive, real-time recognition approach provides a novel solution for precision feeding in group-housed environments and offers valuable insights for improving the efficiency of livestock monitoring and feeding management systems. Full article

Three-Cup Chicken, a traditional Hakka dish, is known for its distinctive umami and salty flavor profile. However, the dynamic evolution of key flavor compounds and associated physicochemical attributes during its characteristic stewing process remains inadequately characterized. Therefore, this study investigated flavor and quality changes in Three-Cup Chicken during stewing using an integrated analytical approach, including gas chromatography-mass spectrometry (GC-MS), gas chromatography-ion mobility spectrometry (GC-IMS), E-tongue, and E-nose, alongside analyses of texture, color, pH, total volatile basic nitrogen (TVB-N), thiobarbituric acid-reactive substances (TBARS), and moisture content. The results revealed that prolonged stewing promoted lipid oxidation, increased hardness, enhanced redness and yellowness, while moisture content gradually decreased. Electronic tongue and nose analyses revealed an increase in saltiness, umami, and sulfur compounds during stewing, complemented by a significant rise in umami amino acids from further analysis. Ten important taste compounds with variable importance in projection (VIP) > 1 and odour activity value (OAV) > 1 were filtered out of 137 volatile compounds, the majority of which were aldehydes. These research findings clearly demonstrate the formation and evolution patterns of the savory and salty flavor profile in Three-Cup Chicken, offering theoretical underpinnings as well as helpful advice for maximizing the dish’s genuine flavor. Full article

Background/Objectives: Ear, nose, and throat (ENT) presentations are common across the UK healthcare system and are often managed initially by junior doctors on call. Short, structured teaching interventions (e.g., boot camps and simulation workshops) have been introduced to improve confidence and preparedness. This review evaluated evidence published since 2015 on such ENT teaching interventions for junior doctors, examining effectiveness, study design, and outcome measures. Methods: Five databases were searched (January 2015–July 2025). Eligible studies assessed ENT-specific courses for junior doctors and reported outcomes on confidence, preparedness, knowledge, or performance. Study quality was appraised using the Medical Education Research Study Instrument (MERSQI). Owing to heterogeneity, findings were narratively synthesised in line with Synthesis Without Meta-analysis (SWiM) guidance. Results: Eleven studies (n = 591) met inclusion criteria: nine single-group pre–post studies, one two-group comparative study, and one randomised controlled trial (RCT). Most studies reported increased confidence after the interventions, while three also showed gains in knowledge. A minority reported improvement using blinded performance assessments. Overall methodological quality assessed using MERSQI scores was moderate (mean 10.0/18). Limitations included reliance on self-reported outcomes, limited use of control groups, and generally short follow-up periods. Conclusions: Short, structured ENT courses for junior doctors are associated with immediate improvements in confidence and knowledge, with some evidence of objective performance gains. However, the predominance of single-arm designs and brief follow-up limits causal inference and conclusions regarding retention, workplace behaviour, and patient outcomes. More robust comparative studies with blinded assessment and longitudinal follow-up are needed to determine sustained impact. Full article

Nanotechnology offers powerful new tools to enhance food quality monitoring and safety assurance. In the food industry, nanoscale materials (e.g., metal, metal oxide, carbon, and polymeric nanomaterials) are being integrated into sensory systems to detect spoilage, contamination, and intentional food tampering with unprecedented sensitivity. Nanosensors can rapidly identify foodborne pathogens, toxins, and chemical changes that signal spoilage, overcoming the limitations of conventional assays that are often slow, costly, or require expert operation. These advances translate into improved food safety and extended shelf-life by allowing early intervention (for example, via antimicrobial nano-coatings) to prevent spoilage. This review provides a comprehensive overview of the types of nanomaterials used in food sensory applications and their mechanisms of action. We examine current applications in detecting food spoilage indicators and adulterants, as well as recent innovations in smart packaging and continuous freshness monitoring. The advantages of nanomaterials—including heightened analytical sensitivity, specificity, and the ability to combine sensing with active preservative functions—are highlighted alongside important toxicological and regulatory considerations. Overall, nanomaterials are driving the development of smarter food packaging and sensor systems that promise safer foods, reduced waste, and empowered consumers. However, realizing this potential will require addressing safety concerns and establishing clear regulations to ensure responsible deployment of nano-enabled food sensing technologies. Representative figures of merit include Au/AgNP melamine tests with LOD 0.04–0.07 mg L⁻¹ and minute-scale readout, a smartphone Au@carbon-QD assay with LOD 3.6 nM, Fe₃O₄/DPV detection of Sudan I at 0.001 µM (linear 0.01–20 µM), and a reusable Au–Fe₃O₄ piezo-electrochemical immunosensor for aflatoxin B1 with LOD 0.07 ng mL⁻¹ (≈15 × reuse), alongside freshness labels that track TVB-N/amine in near-real time and e-nose arrays distinguishing spoilage stages. Full article

We consider the flow of a viscous fluid (power-law, non-Newtonian) injected into a gap of height H between two horizontal plates. When the viscosity of the ambient (displaced) fluid is negligible, the injected fluid forms a tail-slug in contact with both plates connected (at a moving grounding line) to a leading gravity current (GC) whose interface does not touch the top of the gap. Surface tension menisci may appear at the grounding line and nose of the GC. Such systems, of interest in the injection molding industry, have been investigated recently in the framework of the lubrication theory for the volume $\mathcal{V}=q{t}^{\alpha }$ (q and $\alpha$ are positive constants and t is time). Similarity appears for certain values of $\alpha$. The similarity solution of the lubrication model requires manipulations and numerical calculations, which obscure the underlying mechanisms and defy reliable interpretation, because the flow is dependent on four coupled parameters: viscosity exponent n, as well as J, $\sigma$, and ${\sigma }_N$ (the height ratio of the unconfined GC, grounding line meniscus, and nose meniscus to H, respectively). Here we present a significantly simpler box-model analysis, which provides straightforward insights and facilitates the quantitative predictions. Comparisons with the rigorous lubrication-model solution and with previously published data demonstrate that the box model provides a reliable physical description of the system, as well as a fairly accurate prediction of the propagation, for a wide range of parameters. Full article

Background and Objectives: Quantifying nasal proportions is central to facial plastic and reconstructive surgery, yet manual measurements are time-consuming and variable. We sought to develop a simple, reproducible deep learning pipeline that localizes the nose in a single frontal photograph and automatically computes the two-dimensional, photograph-derived apparent nasal index (aNI)—width/height × 100—enabling classification into five standard anthropometric categories. Materials and Methods: From CelebA we curated 29,998 high-quality near-frontal images (training 20,998; validation 5999; test 3001). Nose masks were manually annotated with the VGG Image Annotator and rasterized to binary masks. Ground-truth aNI was computed from the mask’s axis-aligned bounding box. A lightweight one-class YOLOv8n detector was trained to localize the nose; predicted aNI was computed from the detected bounding box. Performance was assessed on the held-out test set using detection coverage and mAP, agreement metrics between detector- and mask-based aNI (MAE, RMSE, R²; Bland–Altman), and five-class classification metrics (accuracy, macro-F1). Results: The detector returned at least one accepted nose box in 3000/3001 test images (99.97% coverage). Agreement with ground truth was strong: MAE 3.04 nasal index units (95% CI 2.95–3.14), RMSE 4.05, and R² 0.819. Bland–Altman analysis showed a small negative bias (−0.40, 95% CI −0.54 to −0.26) with limits of agreement −8.30 to 7.50 (95% CIs −8.54 to −8.05 and 7.25 to 7.74). After excluding out-of-range cases (<40.0), five-class classification on n = 2976 images achieved macro-F1 0.705 (95% CI 0.608–0.772) and 80.7% accuracy; errors were predominantly adjacent-class swaps, consistent with the small aNI error. Additional analyses confirmed strong ordinal agreement (weighted κ = 0.71 linear, 0.78 quadratic; Spearman ρ = 0.76) and near-perfect adjacent-class accuracy (0.999); performance remained stable when thresholds were shifted ±2 NI units and across sex and age subgroups. Conclusions: A compact detector can deliver near-universal nose localization and accurate automatic estimation of the nasal index from a single photograph, enabling reliable five-class categorization without manual measurements. The approach is fast, reproducible, and promising as a calibrated decision-support adjunct for surgical planning, outcomes tracking, and large-scale morphometric research. Full article

Background: Methicillin-resistant Staphylococcus aureus (MRSA) colonization is a risk factor for potential staphylococcal infection and outbreaks. Although it is recommended to obtain a swab culture to detect nasal colonization its necessity in low-prevalence countries is debated. The aim of this study was to determine the prevalence of MRSA nasal colonization, the rate of invasive infection development, and the risk factors for invasive infections in patients admitted to the intensive care unit. Materials and Methods: This retrospective study included patients who were followed up in one of the adult intensive care units at Kayseri City Training and Research Hospital between 1 January 2019 and 31 December 2024 (6 years) and from whom a culture was taken at the time of hospital admission to detect MRSA colonization in the nose. MRSA carriers were examined for the development of any invasive infection caused by MRSA within 28 days of their relevant admission. Results: Over a total period of six years, nasal swab samples were collected from 22,913 patients, and MRSA colonization was detected in 939 (4.0%). Of the patients with MRSA colonization, 32 (3.4%) were excluded from the analysis because they already had invasive MRSA infection. Additionally, 431 patients (45.8%) were excluded from the analysis because they were discharged or died within the first seven days of their admission. Consequently, invasive MRSA infection developed within 28 days in 29 of the 476 patients with MRSA colonization (6.0%). Patients who developed invasive infection had a higher rate of chronic renal failure (p < 0.001), hemodialysis (p < 0.001), central venous catheter (p = 0.028), staying in nursing home (p = 0.001), and a history of hospitalization within the last 90 days (p = 0.015). In the multivariable regression analysis, routine hemodialysis (OR: 5.216, p = 0.015), nursing home stay (OR: 3.668, p = 0.014), and a history of hospitalization within the last 90 days (OR: 2.458, p = 0.028) were found to be risk factors for developing invasive infection. The most common invasive infections were ventilator-associated pneumonia (n = 9), surgical site infection (n = 7), and catheter-related bloodstream infection (n = 6). All 29 strains were susceptible to vancomycin, linezolid, and daptomycin, while one strain was resistant to teicoplanin (3.5%). Conclusions: MRSA colonization has been detected in 4% of patients admitted to the intensive care unit. Screening should be performed because MRSA colonization may be a risk factor for invasive infections; however, screening all patients would be prohibitively expensive and labor-intensive. Instead, it may be more appropriate to identify risk factors and then screen select patients. Full article

Alzheimer’s disease (AD) remains a significant public health challenge, with current treatments limited partly due to the difficulty of delivering therapeutics across the blood–brain barrier (BBB). The nose-to-brain (N-2-B) pathway offers a promising alternative to circumvent the BBB, but no drugs have yet been clinically applied via this route for AD. Mild stress is thought to activate intrinsic protective mechanisms against neurodegeneration, but traditional methods lack specificity and practicality. To address this, we propose the inhalation of mildly heated air as thermal stimulation, which utilizes the N-2-B pathway to induce mild stress and stimulate cerebral activity. This study employs thermal cycling-hyperthermia (TC-HT) in developing thermal cycling-stimulation via nasal inhalation (TCSNI), providing cyclic stimulation to maintain pathway activity while minimizing thermal injury. In C57BL/6 mice, TCSNI showed no adverse olfactory effects. In β-amyloid (Aβ)-treated mice, TCSNI significantly enhanced cognitive performance in Y-maze and novel object recognition (NOR) assessments, suggesting cognitive improvement. Mice hippocampal protein analyses indicated a reduction in Aβ accumulation, alongside increased expression of heat shock protein 70 (HSP70), insulin-degrading enzyme (IDE), and phosphorylated Akt (p-Akt). These results suggest that N-2-B-delivered TCSNI effectively modulates protein expression and enhances cognitive function, highlighting its potential for further exploration in AD treatment. Full article

This study investigated the differences in key volatile organic compounds (VOCs) and flavor characteristics between essential oils (CEOs) from cinnamon bark and leaf. The volatile compounds of essential oils extracted from Cinnamomum cassia (Xijiang) bark (CEOP) and leaf (CEOY) by hydrodistillation were identified using GC-MS. The results showed that the extraction rates of CEOP and CEOY were 1.56% ± 0.02 and 0.83% ± 0.01 (n = 3), respectively. CEOP and CEOY consisted of 45 and 50 compounds, respectively. Odor activity value (OAV) analysis indicated that cinnamaldehyde (OAV = 935), α-caryophyllene (OAV = 77), and borneol (OAV = 4) played key roles in shaping the aroma of CEOP. Meanwhile, cinnamaldehyde (OAV = 849), nerolidol (OAV = 107), and α-caryophyllene (OAV = 58) were the major contributors to the flavor of CEOY. Electronic nose (E-nose) analysis revealed that sensors W5S and W1W were important for detecting aromatic compounds. Sensory evaluation showed that CEOs differed significantly in spicy, floral, and grassy aromas. These differences may be related to the concentrations of compounds such as cinnamaldehyde, α-caryophyllene, and nerolidol, as well as their interactions with olfactory receptors such as OR2W1 and OR1D2. Cinnamaldehyde activates TRPA1 and TRPV1 to elicit the perception of spiciness. Thus, CEOP may be suitable for baked goods, and CEOY may be suitable for ice cream and beverages. In conclusion, this study provides a theoretical foundation for the precise application of CEOs as condiments in food. Full article

Preeclampsia (PE) and intrauterine growth restriction (IUGR) are major pregnancy complications that are linked to placental dysfunction and environmental stimulation such as the use of electronic cigarettes (eCig). This study investigates the molecular impacts of timed eCig exposure in a C57BL/6 mouse model of PE and IUGR using bulk RNA-sequencing of placental tissues. Pregnant mice were exposed to eCig vapor via nose-only system starting at embryonic day 12.5 (eCig-6d, before spiral artery (SA) invasion) or 14.5 (eCig-4d, after SA invasion) until E18.5 (necropsy), with healthy controls exposed to room air (n = 6/group). The eCig-4d group developed PE, whereas the eCig-6d group developed both PE and IUGR. RNA-seq analysis revealed 429 differentially expressed genes (DEGs) in eCig-4d (IUGR-like) group and 64 DEGs in eCig-6d (PE + IUGR-like) group compared to controls. Pathway and gene network analyses indicated that eCig-4d exposure activated NF-κB–driven inflammation, suppressed ECM organization and collagen biosynthesis, and downregulated vasoactive genes/mitochondrial-associated genes (NOS1/2), accompanied by impaired complement initiation and reduced both macrophage and monocyte signals. Similarly, eCig-6d exposure led to downregulation of complement-associated genes and granule-related components, possibly implicating weakened neutrophil responsiveness and compromised inflammatory resolution at the maternal–fetal interface. Our findings align with prior studies on physiological dysfunctions in PE and IUGR, while also providing novel insights into the temporally specific cellular responses induced by eCig exposure. Full article

The study represents the results of a long-term (2016 to 2021) survey on the herpetofauna inhabiting the vicinity of a heavily loaded section of the road E 87. The investigated road splits a Protected site from the net NATURA 2000 BG0000271 “Mandra-Poda”. The Protected site is known for its high biodiversity and its dense populations of vertebrates, which thrive in the area. Directly near the inspected road and on the pavement, we were able to detect five species of snakes, three species of turtles and two species of lizards. Among the squamates, rare observations were made of the European nose-horned viper (Vipera ammodytes), detected twice, and the European glass lizard (Pseudopus apodus), detected three times. Three other species—the Bloched snake (Elaphe sauromates), the Caspian whipsnake (Dolichophis caspius) and the Rhodos green lizard (Lacerta dyplochondrodes)—were found in larger numbers during some of the field surveys and here we provide information concerning the hot moments of their activity in the vicinity of the road. The Grass snakes (Natrix natrix) and the Dice snakes (N. tessellata) formed dense groups in the direct vicinity (closer than one and half meters) of the investigated road section. Despite the high number of recorded snakes and lizards, only isolated cases of vehicle collisions were observed. We suggest that the local squamate population had developed a complex of ethological specifics related to feeding, basking, shading, and copulation, which helped them to benefit from the road and avoid the risks related to the heavy traffic. Full article

Classifying low-concentration Gas Sensor Array (GSA) data is hard due to low SNR, sensor heterogeneity, drift, and small samples. We benchmark time-series-to-image (TS2I) CNNs against time-series classifiers, after reproducing a strong FE-ELM baseline under a shared fold manifest. Using the GSA-LC and GSA-FM datasets, we compare FE-ELM, vector baselines, time-series methods, and TS2I-CNNs with 20 × 5 repeated stratified cross-validation (n = 100). ROCKET delivers the best accuracy on both datasets and is significantly better than TCN and MiniROCKET (paired tests with Holm–Bonferroni, p < 0.05): on GSA-FM, accuracy 0.9721 ± 0.0480 (95% CI [0.9627, 0.9815]) with Macro-F1 0.9757; on GSA-LC, 0.9578 ± 0.0433 (95% CI [0.9493, 0.9663]) with Macro-F1 0.9555. Among image-based models, CNN-RP is the most robust, whereas CNN-GASF lags, especially on GSA-LC. RGB fusion strategies (e.g., with MTF) are dataset-dependent and often inconsistent, and transfer learning with ResNet-18 offers no consistent advantage. Overall, ROCKET ranks first across folds, while CNN-RP is the most reliable TS2I alternative under low-concentration conditions. These results provide a reproducible, fair benchmark for e-nose applications and practical guidance for model selection, while clarifying both the potential and limitations of TS2I. Full article