Spinal cord injury (SCI) impairs motor function and requires rigorous rehabilitative therapy, motivating the development of approaches that are engaging and customizable. Virtual reality (VR) motor training with augmented sensory feedback (ASF) offers a promising pathway to enhance functional outcomes, yet it remains unclear how ASF modalities affect performance and underlying psychophysiological states in persons with SCI. Five participants with chronic incomplete cervical-level SCI controlled a virtual robotic arm with semi-isometric upper-body contractions while undergoing ASF training with either visual feedback (VF) or combined visual plus haptic feedback (VHF). Motor performance (pathlength, completion time), psychophysiological measures (EEG, EMG, EDA, HR), and perceptual ratings (agency, motivation, utility) were assessed before and after ASF training. VF significantly reduced pathlength (-12.5%, = 0.0011) and lowered EMG amplitude (-32.5%, = 0.0063), suggesting the potential for improved motor performance and neuromuscular efficiency. VHF did not significantly improve performance, but trended toward higher cortical engagement. EEG analyses showed VF significantly decreased alpha and beta activity after training, whereas VHF trended toward mild increases. Regression revealed improved performance was significantly ( < 0.05) associated with changes in alpha power, EMG, EDA, and self-reported motivation. ASF type may differentially shape performance and psychophysiological responses in SCI participants. These preliminary findings suggest VR-based ASF as a potent multidimensional tool for personalizing rehabilitation.

Somatosensation is the ability to detect various external and internal stimuli (such as pain, pressure, temperature, or joint position), and its objective and reproducible evaluation is essential for diagnosis, training, and rehabilitation. This study evaluates the methodological quality of two somatosensory assessments in young healthy adults. The tuning fork test (administered on five locations of each hemibody) and the monofilament test (administered on 27 locations of each hemibody, and divided into (i) foot and ankle, (ii) leg and thigh, and (iii) trunk subscales) were applied to 58 students by two raters at three different time points (rater 1 test, rater 1 retest, rater 2 test). The intra- and inter-rater reliability, parallel test reliability, and internal consistency were evaluated for each test and subtest. The tuning fork test showed moderate intra- and inter-rater reliability and good internal consistency. The monofilament test showed good to moderate intra- and inter-rater reliability for foot and ankle locations, but poor intra- and inter-rater reliability for leg, thigh, and trunk locations. The total score, left hemibody score, and right hemibody score of the monofilament test showed good or acceptable consistency with leg and thigh subscales, but poor or unacceptable consistency with foot, ankle, and trunk subscales. No acceptable parallel test reliabilities were found between the tuning fork test and the monofilament test. The tuning fork test is a reliable assessment of deep somatosensory function in the lower extremities of healthy young adults. The commercially available monofilament test kits are sufficient to investigate the superficial somatosensitivity of feet and ankles, but are insufficient for an objective evaluation of leg, thigh, and trunk regions.

Accurate detection of cardiac arrhythmias from electrocardiogram (ECG) signals is essential for the early diagnosis of cardiovascular diseases but remains challenging due to the complex, non-linear nature of ECG waveforms. This study proposes HCTG-Net, a Hybrid CNN-Transformer Network with Gated Fusion, designed to jointly capture local morphological features and long-range temporal dependencies in ECG data. The model employs a dual-branch architecture, where a residual CNN extracts localized waveform patterns and a Transformer branch models global temporal context. A learnable gated fusion mechanism adaptively balances and integrates features from both branches at the per-dimension level. Experiments conducted on the MIT-BIH Arrhythmia Database demonstrate that HCTG-Net achieves superior performance compared with existing methods, reaching an overall accuracy of 0.9946 and F1-score of 0.9711. Visualization results show well-clustered feature distributions, confirming robust feature learning, while ablation studies verify the complementary roles of the CNN, Transformer, and fusion modules. Overall, HCTG-Net offers a powerful and adaptive framework for automatic ECG-based arrhythmia diagnosis and holds strong potential for real-time clinical and wearable healthcare applications.

: Cone-beam computed tomography (CBCT) was used for a 1-year follow-up of a randomized clinical trial to compare a stainless-steel Tornado file system with OneShape and WaveOne rotary systems for biomechanical canal preparation, as indicated by radiolucency sizes of periapical lesions. : Lower molars with necrotic pulps and periapical lesions were randomly divided into three groups (n = 20) according to three rotary file systems. After root canal treatment, clinical and assessment of the CBCT periapical index scores were blindly evaluated at one year using pre- and post-instrumentation CBCT images. Statistical analysis was performed to compare the three systems at a -value of 0.05. : The results revealed a significant decrease in the size of apical radiolucency in each group after one-year follow-up, with no statistically significant difference among the three systems ( > 0.05). : CBCT is a valuable biomedical imaging modality for assessing periapical lesion healing. Tornado, WaveOne, and OneShape systems can be used with similar efficacy for root canal preparation in teeth with periapical lesions. : The study was retrospectively registered with ClinicalTrials.gov (NCT06752837). Date of Registration: 30 December 2024. The CONSORT group has identified it as essential.

Biomedical ontologies are critical for structuring domain knowledge and enabling integrative analyses in the life sciences. Traditional ontology development is labor-intensive, requiring extensive expert curation. Recent advances in artificial intelligence, particularly large language models (LLMs), have opened new possibilities to automate and enhance various aspects of bio-ontology research. This review article synthesizes findings from recent studies on LLM-assisted ontology creation, mapping, integration, and semantic search, while addressing challenges such as bias, reliability, and ethical concerns. We also discuss promising future directions and emerging trends that may further transform the way biomedical ontologies are developed, maintained, and used.

Multiple imputation (MI) is widely used for handling missing data. However, applying penalized methods after MI can be challenging because variable selection may be inconsistent across imputations. We propose a two-step variable selection method for multiply imputed datasets with survival outcomes: apply LASSO or ALASSO to each MI dataset, followed by ridge regression, and combine estimates using variable selected in any or d% (d = 50, 70, 90, 100) of the MI datasets. For comparison, we also fit stacked MI datasets with weighted penalized regression and a group LASSO approach that enforces consistent selection across imputations. Simulations with Cox models evaluated tuning by AIC, BIC, cross-validation at the minimum error, and the 1SE rule. Across scenarios, performance differed by both the penalization and the selection rule. More conservative choices such as ALASSO with BIC and a 50% inclusion frequency tended to control false positive and gave more stable calibration. The grouped approach achieved comparable selection with modestly higher estimation error. Overall, no single method consistently outperformed others across all scenarios. Our findings suggest that practitioners should weigh trade-offs between selection stability, estimation accuracy, and calibration when applying penalized methods to multiply imputed survival data.

The green synthesis of silver nanoparticles (SNPs) using medicinal plants provides a sustainable and eco-friendly approach to nanoparticle production with promising biomedical potential. In this study, and aqueous leaf extracts were employed as reducing and stabilizing agents to synthesize SNPs (RcSNPs) and SNPs (AbSNPs). The nanoparticles were characterized using ultraviolet-visible spectroscopy, dynamic light scattering, Fourier-transform infrared spectroscopy, scanning electron microscopy, transmission electron microscopy, thermogravimetric analysis, and differential scanning calorimetry to evaluate their physicochemical and thermal properties. RcSNPs and AbSNPs were predominantly spherical, with average sizes of 15-20 nm and 23-28 nm, respectively, and exhibited stability up to ~90 °C. Biological evaluations demonstrated potent antimicrobial, antioxidant, anti-inflammatory, anti-tyrosinase, and cytotoxic activities. Notably, RcSNPs and AbSNPs induced apoptosis through mitochondrial pathway modulation and showed superior cytotoxicity compared to crude plant extracts and several previously reported SNPs. These findings indicate that phytochemical-mediated SNPs not only provide a green route of synthesis but also exhibit multifunctional bioactivities, which may support their potential applications as antimicrobial, antioxidant, depigmenting, and anticancer agents in biomedical and pharmaceutical fields.

: Extreme lateral interbody fusion (XLIF) is a minimally invasive spine procedure that traditionally relies on fluoroscopy and neuromonitoring for safe disc space access and instrumentation. Augmented reality (AR) navigation offers real-time anatomical visualization and may reduce fluoroscopy use. This is the first description of applying augmented reality to lateral spine surgery. : We conducted a case series of five patients who underwent AR-guided LLIF between May 2024 and July 2025. Surgery was performed in either lateral decubitus or prone transpsoas (PTP) orientation. AR navigation was performed using the Augmedics xvision Spine System, with intraoperative CT-based registration and optical tool tracking. Clinical and operative data, including operative time, estimated blood loss (EBL), length of stay (LOS), radiation exposure, instrumentation accuracy, and postoperative outcomes, were collected and analyzed. : Five patients (4 female, 1 male; age > 65; BMI range 20.7-37.2) underwent AR-guided XLIF across 8 levels (L2-L5). The mean operative time was 5 h 1 min (range: 2 h 8 min-6 h 45 min), and mean EBL was 94 mL. Mean LOS was 5.85 days (range: 2-10). Mean radiation exposure was 21.73 mGy, significantly lower than published averages for fluoroscopy-guided XLIF (108.6 mGy). At follow-up, all patients reported pain reduction, with 4/5 achieving complete symptom resolution. Instrumentation accuracy was confirmed radiographically in all cases. : This clinical series demonstrates the first clinical application of AR to lateral lumbar interbody fusion. AR navigation was feasible, safe, and effective, providing accurate disc space access and instrumentation with markedly reduced radiation exposure. These findings support AR as a promising adjunct to improve safety, efficiency, and workflow in lateral spine surgery.

Ischemic heart disease (IHD) is the leading cause of death worldwide. Although 90% of the intramyocardial blood volume resides in the microvasculature, clinical imaging methods cannot visualize the microvascular coronary network in vivo, and non-invasive hemodynamic estimates overlook patient-specific microcirculatory contributions. Herein, we present a multiscale framework to extend the epicardial coronary tree and generate 1D microvascular networks in the myocardium based on ferumoxytol-enhanced magnetic resonance coronary imaging and fractional myocardial blood volume (fMBV) maps. Synthetic arterial networks were constructed from MRI data belonging to three swine, four healthy volunteers, and one IHD patient using a modified multistage, adaptive constrained constructive optimization approach. Hemodynamic simulations were performed in synthetic arterial networks. Morphological parameters were compared with empirical models. In 126 arterial networks ( = 6000 terminal segments per subject per seed; six seeds per coronary vessel), the morphometry was strongly correlated with empirical data ( > 0.87), with low variability ( < 0.01) across multiple rounds of network simulations. Mixed-effects models and a Dynamic Time Warping analysis confirmed robustness and repeatability. In the IHD patient, simulated arterial networks ( = 15) reproduced tissue-dependent morphological and functional signatures consistent with coronary autoregulation in scar and hypoperfused tissues. The findings establish an early potential for patient-specific microvascular network synthesis and hemodynamic simulations from MRI data.

Phonoenterography (PEG) offers a non-invasive and radiation-free technique to assess gastrointestinal activity through acoustic signal analysis. In this feasibility study, 110 high-resolution PEG recordings (44.1 kHz, 16-bit) were acquired from eight healthy individuals, yielding 6314 prominent bowel sound (PBS) segments through automated segmentation. Each event was characterized using a 279-feature acoustic profile comprising Mel-frequency cepstral coefficients (MFCCs), their first-order derivatives (Δ-MFCCs), and six global spectral parameters. After normalization and dimensionality reduction with PCA and UMAP (cosine distance, 35 neighbors, minimum distance = 0.01), five clustering strategies were evaluated. K-Means (k = 5) achieved the most favorable balance between cluster quality (silhouette = 0.60; Calinski-Harabasz = 19,165; Davies-Bouldin = 0.68) and interpretability, consistently identifying five acoustic patterns: single-burst, multiple-burst, harmonic, random-continuous, and multi-modal. Temporal modeling of clustered events further revealed distinct sequential dynamics, with Single-Burst events showing the longest dwell times, random continuous the shortest, and strong diagonal elements in the transition matrix confirming measurable state persistence. Frequent transitions between random continuous and multi-modal states suggested dynamic exchanges between transient and overlapping motility patterns. Together, these findings demonstrate that unsupervised PEG-based analysis can capture both acoustic variability and temporal organization of bowel sounds. This annotation-free approach provides a scalable framework for real-time gastrointestinal monitoring and holds potential for clinical translation in conditions such as postoperative ileus, bowel obstruction, irritable bowel syndrome, and inflammatory bowel disease.

This paper explores the development of a Smart e-Textile Singlet designed to enhance geriatric care through continuous monitoring of vital health parameters. The proposed garment integrates various sensors to measure core body temperature, blood oxygen saturation, respiration rate, blood pressure, pulse, electrocardiogram (ECG), activity level, and risk of falls. Leveraging advanced technologies such as inertial measurement unit (IMU) sensors, thermoelectric materials, and piezoelectric fibers, the e-textile ensures both functionality and sustainability. Additionally, artificial intelligence algorithms are employed to provide near-real-time feedback and early warnings, significantly improving health management for elderly individuals. This innovative approach not only promotes autonomy and well-being among the elderly but also alleviates the workload of healthcare providers. The Smart e-Textile Singlet represents a multi-sensor solution by offering a holistic monitoring system.

The rising frequency of heat-related illnesses (HRIs) under climate change presents urgent public health challenges, particularly in urban environments. This study develops an explainable machine learning (ML) model to predict HRI risk using metrological data from seven major South Korean metropolitan cities between May and September 2021-2024. We applied eXtreme Gradient Boosting (XGBoost) to model relationships between daily meteorological variables, including maximum and mean daily temperatures, humidity, solar radiation, wind speed, and precipitation, and HRI occurrence. Model performance was validated using 2025 data and demonstrated strong predictive accuracy, with area under the curve (AUC) values 0.895. To enhance interpretability, Shapley Additive exPlanations (SHAP) analysis identified mean daily temperature, solar radiation, and minimum temperature as the strongest contributors to HRI risk. Time-series comparisons of predicted and actual HRI occurrences further validated the model's effectiveness in real-world settings. These findings underscore the potential of eXplainable Artificial Intelligence (XAI) for localized health-risk forecasting and support a data-driven basis for developing early warning systems for climate-sensitive diseases to guide proactive public health planning amid escalating urban heat risks.

Although AI-mediated approaches provide promising support for bioengineering using training datasets, their application in microalgal research remains limited. In this study, ChatGPT-4.0, an easily accessible AI model, was employed to optimize culture conditions and evaluate the industrial potential of the isolated diatom . Culture optimization was conducted using response surface methodology, in which pH, temperature, and salinity were selected as independent variables. ChatGPT assisted in determining the design and suggested a face-centered central composite design. The optimal conditions for biomass production were determined to be pH 8.30, 23 °C, and 34.24 psu. Analysis of variance revealed significant quadratic effects ( < 0.05), indicating curvature in the response surface. Fatty acid profiling showed high levels of palmitoleic acid, palmitic acid, and eicosapentaenoic acid. Pigment analysis further indicated a high abundance of fucoxanthin, diadinoxanthin, and diatoxanthin. Based on the analyzed compounds, ChatGPT suggested potential applications of the algal strain across various industrial sectors. The most relevant application was identified as aquafeed, as the strain contains metabolites known to enhance pigmentation, growth, and immune responses in aquaculture species. Overall, this study demonstrates ChatGPT-mediated bioengineering as a practical strategy for optimizing culture conditions and evaluating the industrial potential of novel microalgal strains.

Research in the biomedical field often faces challenges due to the scarcity and high cost of data, which significantly limit the development and application of machine learning models. This paper introduces a data-centric AI framework for EEG-based emotion recognition that emphasizes improving data quality rather than model complexity. Instead of proposing a deep architecture, we demonstrate how participant-guided noise filtering combined with systematic data augmentation can substantially enhance system performance across multiple classification settings: binary (high vs. low arousal), four-quadrant emotions, and seven discrete emotions. Using the SEED-VII dataset, we show that these strategies consistently improve accuracy and F1 scores, achieving competitive or superior performance compared to more sophisticated published models. The findings highlight a practical and reproducible pathway for advancing biomedical AI systems, showing that prioritizing data quality over architectural novelty yields robust and generalizable improvements in emotion recognition.

Collagen peptides are key structural proteins that play an important role in maintaining the integrity and proper function of multiple tissues in the human body. Their breakdown is recognized as an important biomarker for various degenerative conditions, including the loss of muscle mass, joint and bone disorders, and compromised skin health. Current analytical approaches for collagen detection, such as ultraviolet spectrometry, enzyme-linked immunosorbent assay (ELISA), high-performance liquid chromatography (HPLC), and histochemical staining, are widely used but often expensive, time-consuming, and reliant on specific laboratory instrumentation, limiting their practicality for routine or rapid diagnostics. This study reports a novel biosensor for collagen peptide detection based on molecularly imprinted polymers (MIPs) integrated with screen-printed electrodes (SPEs). Electrochemical measurements revealed a clear correlation between collagen concentration and current response, confirming effective molecular binding within the imprinted matrix. The optimized MIP-modified electrode exhibited a detection range of 0.1-1000 µg/mL with a limit of detection (LOD) of 1.0106 µg/mL, limit of quantification (LOQ) of 4.46 µg/mL, sensitivity of 8.3816, and correlation coefficient (R = 0.9436). These results highlight strong selectivity and sensitivity toward collagen peptides. The proposed MIP-based biosensor provides a rapid, low-cost platform for detecting collagen degradation products and holds potential for early diagnosis and future clinical applications in degenerative disease monitoring.

The long-term stability of bioactive dental cements in acidic environments is not yet fully understood, despite their extensive clinical use in restorative and endodontic procedures. The objective of this study is to evaluate the degradation behaviour and mechanical stability of one glass ionomer cement (GC FUJI IX) and two calcium-silicate-based materials (Biodentine and Biodentine XP 500) under simulated acidic oral conditions. A total of 18 samples were prepared and distributed into three groups. The materials were immersed in a solution with a pH of 4.5, and their performance was assessed through a number of different methods. These included mass-loss measurements, corrosion-rate calculations, Vickers microhardness testing, and SEM to characterise the surfaces. Biodentine exhibited the highest degradation, followed by Bio-Dentine XP 500 and GC FUJI IX. The data were confirmed by one-way ANOVA and a post hoc Tukey's test. This indicated a statistically significant superiority ( < 0.05) of Biodentine XP 500 over glass ionomers in terms of surface hardness maintenance under acidic conditions. Biodentine, a calcium silicate-based material, demonstrated inferior chemical stability compared to GC FUJI IX and Biodentine XP 500, likely due to its modified calcium-silicate formulation that limits ionic dissolution. In addition, the study revealed that Biodentine XP 500 exhibited the highest Vickers hardness under acidic conditions. The findings reported in this study offer valuable insights into the material selection process for low-pH clinical scenarios and contribute to a more comprehensive understanding of the chemical-mechanical stability of modern bioactive dental restoratives.

(1) Current synthetic small-diameter vascular grafts fail frequently due to anastomotic hyperplasia and thrombosis caused by mechanical mismatch and incomplete reendothelialization. Polydioxanone near-field electrospun (NFES) vascular templates feature programmable pore sizes to facilitate transmural ingrowth of endothelial cells and show promise in reducing mechanical mismatch, but their potential as drug delivery systems remains unexplored. It was hypothesized that Manuka honey incorporation in NFES templates could reduce neutrophil extracellular trap (NET) release but decrease mechanical strength. (2) Templates were fabricated using 90 mg/mL polydioxanone in 1,1,1,3,3,3-hexafluoro-2-propanol (HFP) and Manuka honey concentrations of 0%, 0.1%, 1%, and 10% /. Wall thickness (197-236 μm), mechanical properties, Manuka honey elution, and NET release were quantified. (3) The 0.1% and 1% templates best mimicked native vessel mechanics, outperforming the pure HFP template in tensile strength and burst pressure. The 10% templates exhibited significant mechanical strength reductions. Manuka honey elution exhibited a burst release within the first three hours, and all honey was eluted by day three. NET release was elevated in 10% and control groups but was not significantly different from 0.1% and 1%. (4) Overall, low concentrations of Manuka honey maintained mechanical compatibility, but elution must be optimized for immunomodulation, rejecting the initial hypothesis.

(1) Background: The increasing utilization of three-dimensional (3D) photogrammetry has elevated craniofacial analysis to new dimensions. This scoping review seeks to provide a comprehensive overview of the current applications of 3D photogrammetry-supported craniofacial analysis within orthodontic practice, assess its technical superiority, and explore potential areas for enhancement. (2) Methods: A comprehensive search of the literature was carried across three electronic databases (PubMed, Web of Science, Embase). Two independent reviewers screened the articles and extracted data in accordance with the PRISMA-ScR guideline. The primary findings from the included articles were synthesized and analyzed qualitatively. (3) Results: A total of 479 records were obtained initially, with 53 articles ultimately included after removing duplicates and applying eligibility criteria. The application of 3D photogrammetry in craniofacial analysis has become prevalent in orthodontic practice, encompassing normative facial anthropometry, orthodontic problem finding, orthodontic treatment optimization, and treatment outcome evaluation. (4) Conclusion: 3D photogrammetry offers orthodontists a precise and efficient imaging technique for craniofacial analysis.

Acute low back pain can impair motor control, yet its effects on force steadiness and cortical activity remain unclear.

The global rise in cancer incidence and mortality represents a major challenge for modern healthcare. Although current screening programs rely mainly on histological or immunological biomarkers, cancer is a multifactorial disease in which biological, psychological, and behavioural determinants interact. Psychological dimensions such as stress, anxiety, and depression may influence vulnerability and disease evolution through neuro-endocrine, immune, and behavioural pathways, especially by affecting adherence to therapeutic recommendations. However, these dimensions remain underexplored in current screening workflows. This review synthesizes current evidence on the integration of biological markers (tumor and inflammatory biomarkers), psychometric profiling (stress, depression, anxiety, personality traits), and behavioural digital phenotyping (facial micro-expressions, vocal tone, gait/posture metrics) for potential early cancer risk evaluation. We examine recent advances in computational sciences and artificial intelligence that could enable multimodal signal harmonization, structured representation, and hybrid data fusion models. We discuss how structured computational information management may improve interpretability and may support future AI-assisted screening paradigms. Finally, we highlight the relevance of digital health infrastructure and telemedical platforms in strengthening accessibility, continuity of monitoring, and population-level screening coverage. Further empirical research is required to determine the true predictive contribution of psychological and behavioural modalities beyond established biological markers.

Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized by motor and non-motor symptoms, among which vocal impairment is one of the earliest and most prevalent. In recent years, voice analysis supported by machine learning (ML) and deep learning (DL) has emerged as a promising non-invasive method for early PD detection. We conducted a systematic review searching PubMed, Scopus, IEEE Xplore, and Web of Science databases for studies published between 2020 and September 2025. A total of 69 studies met the inclusion criteria and were analyzed in terms of dataset characteristics, speech tasks, feature extraction techniques, model architectures, validation strategies, and performance outcomes. Classical ML models such as Support Vector Machines (SVMs) and Random Forests (RFs) achieved high accuracy on small, homogeneous datasets, while DL architectures, particularly Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Transformer-based foundation models, demonstrated greater robustness and scalability across languages and recording conditions. Despite these advances, persistent challenges such as dataset heterogeneity, class imbalance, and inconsistent validation practices continue to hinder reproducibility and clinical translation. Overall, the field is transitioning from handcrafted feature-based pipelines toward self-supervised, representation-learning frameworks that promise improved generalizability. Future progress will depend on the development of large, multilingual, and openly accessible datasets, standardized evaluation protocols, and interpretable AI frameworks to ensure clinically reliable and equitable voice-based PD diagnostics.