Search Results (270)

Affective computing has emerged as a pivotal field in human–computer interaction. Recognizing human emotions through electroencephalogram (EEG) signals can advance our understanding of cognition and support healthcare. This study introduces a novel subject-independent emotion recognition framework by integrating multiple EEG emotion databases (DEAP, MAHNOB HCI-Tagging, DREAMER, AMIGOS and REFED) into a unified dataset. EEG segments were transformed into feature vectors capturing statistical, spectral, and entropy-based measures. Standardized pre-processing, analysis of variance (ANOVA) F-test feature selection, and six machine learning models were applied to the extracted features. Classification models such as Decision Tree, Discriminant Analysis, Support Vector Machine (SVM), K-Nearest Neighbor (K-NN), Naive Bayes, and Artificial Neural Networks (ANN) were considered. Experimental results demonstrate that SVM achieved the best performance for arousal classification (70.43%), while ANN achieved the highest accuracy for valence classification (68.07%), with both models exhibiting strong generalization across subjects. The results highlight the feasibility of developing biomimetic brain–computer interface (BCI) systems for objective assessment of emotional intelligence and its cognitive underpinnings, enabling scalable applications in affective computing and adaptive human–machine interaction. Full article

In this study, we describe a non-invasive approach to objectively assess fragrance-induced emotions using multiplex salivary biomarker profiling. Traditional self-reports, physiological monitoring, and neuroimaging remain limited by subjectivity, invasiveness, or poor temporal resolution. Saliva offers an advantageous alternative, reflecting rapid neuroendocrine changes linked to emotional states. We combined four key salivary biomarkers, cortisol, alpha-amylase, dehydroepiandrosterone, and oxytocin, to capture multidimensional emotional responses. Two clinical studies (n = 30, n = 63) and one user study (n = 80) exposed volunteers to six fragrances, with saliva collected before and 5 and 20 min after olfactory stimulation. Subjective emotional ratings were also obtained through questionnaires or an implicit approach. Rigorous analytical validation accounted for circadian variation and sample stability. Biomarker patterns revealed fragrance-induced emotional profiles, highlighting subgroups of participants whose biomarker dynamics correlated with particular emotional states. Increased oxytocin and decreased cortisol levels aligned with happiness and relaxation; in comparison, distinct biomarker combinations were associated with confidence or dynamism. Classification and Regression Trees (CART) analysis results demonstrated high sensitivity for detecting these profiles. Validation in an independent cohort using an implicit association test confirmed concordance between molecular profiles and behavioral measures, underscoring the robustness of this method. Our findings establish salivary biomarker profiling as an objective tool for decoding real-time emotional responses. Beyond advancing affective neuroscience, this approach holds translational potential in personalized fragrance design, sensory marketing, and therapeutic applications for stress-related disorders. Full article

Background/Objectives: Electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) can objectively capture the spatiotemporal dynamics of brain activity during affective cognition, and their combination is promising for improving emotion recognition. However, multi-modal cross-subject emotion recognition remains challenging due to heterogeneous signal characteristics that hinder effective fusion and substantial inter-subject variability that degrades generalization to unseen subjects. Methods: To address these issues, this paper proposes DC-AGIN, a dual-contrastive learning attention graph isomorphism network for EEG–fNIRS emotion recognition. DC-AGIN employs an attention-weighted AGIN encoder to adaptively emphasize informative brain-region topology while suppressing redundant connectivity noise. For cross-modal fusion, a cross-modal contrastive learning module projects EEG and fNIRS representations into a shared latent semantic space, promoting semantic alignment and complementarity across modalities. Results: To further enhance cross-subject generalization, a supervised contrastive learning mechanism is introduced to explicitly mitigate subject-specific identity information and encourage subject-invariant affective representations. Experiments on a self-collected dataset are conducted under both subject-dependent five-fold cross-validation and subject-independent leave-one-subject-out (LOSO) protocols. The proposed method achieves 96.98% accuracy in four-class classification in the subject-dependent setting and 62.56% under LOSO. Compared with existing models, DC-AGIN achieves SOTA performance. Conclusions: These results demonstrate that the work on attention aggregation, cross-modal and cross-subject contrastive learning enables more robust EEG-fNIRS emotion recognition, thus supporting the effectiveness of DC-AGIN in generalizable emotion representation learning. Full article

Non-invasive electrocardiogram (ECG) and respiratory signals are easy to acquire via low-cost sensors, making them promising alternatives for sleep staging. However, existing methods using these signals often yield insufficient accuracy. To address this challenge, we incrementally optimized the sleep staging model by designing a structured experimental workflow: we first preprocessed respiratory and ECG signals, then extracted fused features using an enhanced feature selection technique, which not only reduces redundant features, but also significantly improves the class discriminability of features. The resulting fused features serve as a reliable feature subset for the classifier. In the meantime, we proposed a hybrid optimization algorithm that integrates the snake optimization algorithm (SO) and ant colony optimization algorithm (ACO) for automated hyperparameter optimization of support vector machines (SVMs). Experiments were conducted using two PSG-derived public datasets, the Sleep Heart Health Study (SHHS) and MIT-BIH Polysomnography Database (MIT-BPD), to evaluate the classification performance of multimodal features compared with single-modal features. Results demonstrate that the bimodal staging using SHHS multimodal signals significantly outperformed single-modal ECG-based methods, and the overall accuracy of the SHHS dataset was improved by 12%. The SVM model optimized using the hybrid Snake–ACO algorithm achieved an average accuracy of 89.6% for wake versus sleep classification on the SHHS dataset, representing a 5.1% improvement over traditional grid search methods. Under the subject-independent partitioning experiment, the wake versus sleep classification task maintained good stability with only a 1.8% reduction in accuracy. This study provides novel insights for non-invasive sleep monitoring and clinical decision support. Full article

Gaming Disorder (GD) is becoming more widely acknowledged as a behavioral addiction characterized by impaired control and functional impairment. While resting-state impairments are well understood, the neurophysiological dynamics during active gameplay remain underexplored. This study identified task-based occipital EEG biomarkers of GD and assessed their diagnostic utility. Occipital EEG (O1/O2) data from 30 participants (15 with GD, 15 controls) collected during active mobile gaming were used in this study. Spectral, temporal, and nonlinear complexity features were extracted. Feature relevance was ranked using Random Forest, and classification performance was evaluated using Leave-One-Subject-Out (LOSO) cross-validation to ensure subject-independent generalization across five models (Random Forest, KNN, SVM, Decision Tree, ANN). The GD group exhibited paradoxical “spectral slowing” during gameplay, characterized by increased Delta/Theta power and decreased Beta activity relative to controls. Beta variability was identified as a key biomarker, reflecting altered attentional stability, while elevated Alpha power suggested potential neural habituation or sensory gating. The Decision Tree classifier emerged as the most robust model, achieving a classification accuracy of 80.0%. Results suggest distinct neurophysiological patterns in GD, where increased low-frequency power may reflect automatized processing or “Neural Efficiency” despite active task engagement. These findings highlight the potential of occipital biomarkers as accessible and objective screening metrics for Gaming Disorder. Full article

Background/Objectives: Parkinsonian gait is characterized by impaired motor adaptability, altered temporal organization, and reduced movement variability. While data augmentation is commonly used to mitigate class imbalance in gait-based machine learning models, conventional strategies often ignore physiological differences between healthy and pathological movements, potentially distorting meaningful motor dynamics. This study explores whether preserving healthy motor variability while selectively augmenting pathological gait signals can improve the robustness and physiological coherence of gait pattern classification models. Methods: Eight patients with Parkinsonian gait patterns and forty-eight healthy participants performed walking tasks on the Motigravity platform under hypogravity conditions. Full-body kinematic data were acquired using wearable inertial sensors. A selective augmentation strategy based on smooth time-warping was applied exclusively to pathological gait segments (×5, σ = 0.2), while healthy gait signals were left unaltered to preserve natural motor variability. Model performance was evaluated using a hybrid convolutional neural network–long short-term memory (CNN–LSTM) architecture across multiple augmentation configurations. Results: Selective augmentation of pathological gait signals achieved the highest classification performance (94.1% accuracy, AUC = 0.97), with balanced sensitivity (93.8%) and specificity (94.3%). Performance decreased when augmentation exceeded an optimal range of variability, suggesting that beneficial augmentation is constrained by physiologically plausible temporal dynamics. Conclusions: These findings demonstrate that physiology-informed, selective data augmentation can improve gait pattern classification under constrained data conditions. Rather than supporting disease-specific diagnosis, this proof-of-concept study highlights the importance of respecting intrinsic differences in motor variability when designing augmentation strategies for clinical gait analysis. Future studies incorporating disease-control cohorts and subject-independent validation are required to assess specificity and clinical generalizability. Full article

Background/Objectives: Childhood obesity remains a major public health issue, particularly in Mediterranean countries such as Greece. Although parental influences on children’s weight have been extensively studied, fewer studies have jointly examined parental mental health, feeding practices, sociodemographic factors, and biological stress markers. This study aimed to investigate associations between psychological status, educational level, feeding behaviors, and children’s Body Mass Index (BMI) in a Greek sample. A pilot assessment of salivary cortisol was included in evaluating its feasibility as an objective biomarker of parental stress. Subjects and Methods: A total of 103 parent–child dyads participated in this cross-sectional study. Children’s BMI was classified using World Health Organization (WHO) growth standards. Parental stress, anxiety, and depressive symptoms were assessed using the Perceived Stress Scale-14 (PSS-14) and the Depression Anxiety Stress Scale-21 (DASS-21) questionnaires. Feeding practices were evaluated with the Comprehensive Feeding Practices Questionnaire (CFPQ). Statistical analyses included Pearson correlations, independent samples t-tests, one-way ANOVA, Mann–Whitney U, and Kruskal–Wallis tests. A subsample provided saliva samples for cortisol analysis to assess feasibility and explore the potential associations with parental stress indicators. Results: Parental BMI showed a strong positive association with child BMI (p = 0.002). Higher parental anxiety (p = 0.002) and depression (p = 0.009) were also associated with increased child BMI. Restrictive (p < 0.001) and emotion-driven (p < 0.001) feeding practices were associated with higher child BMI, whereas monitoring (p = 0.013) and health-promoting feeding practices (p = 0.001) appeared protective. Lower parental education was related to a higher BMI in both parents (p = 0.001) and children (p = 0.002) and to more frequent use of restrictive feeding strategies (p = 0.001). WHO charts identified a greater proportion of children as overweight or obese compared with the Centers for Disease Control and Prevention (CDC) criteria. The analysis showed statistically significant differences between the two classification systems (χ² (4) = 159.704, p < 0.001), indicating that BMI categorization varies considerably depending on the reference system used. No significant associations were observed with residential environment or salivary cortisol, likely due to the limited size of the pilot biomarker subsample. Conclusions: The findings highlight the combined effect of parental mental health status, educational level, and feeding practices on child BMI within the Greek context. The preliminary inclusion of a biological stress marker provides added value to the existing research in this area. These results underscore the importance of prevention strategies that promote parental psychological wellbeing and responsive feeding practices while addressing socioeconomic disparities to reduce the childhood obesity risk. Full article

With the ongoing miniaturization of smart devices, fine-grained hand gesture recognition using millimeter-wave radar has attracted increasing attention, yet practical deployment remains challenging in continuous-gesture segmentation, robust feature extraction, and reliable classification. This paper presents an end-to-end fine-grained gesture recognition framework based on frequency modulated continuous wave(FMCW) millimeter-wave radar, including gesture design, data acquisition, feature construction, and neural network-based classification. Ten gesture types are recorded (eight valid gestures and two return-to-neutral gestures); for classification, the two return-to-neutral gesture types are merged into a single invalid class, yielding a nine-class task. A sliding-window segmentation method is developed using short-time Fourier transformation(STFT)-based Doppler-time representations, and a dataset of 4050 labeled samples is collected. Multiple signal classification(MUSIC)-based super-resolution estimation is adopted to construct range–time and angle–time representations, and instance-wise normalization is applied to Doppler and range features to mitigate inter-individual variability without test leakage. For recognition, a variable-channel deep residual shrinkage network (DRSN) is employed to improve robustness to noise, supporting single-, dual-, and triple-channel feature inputs. Results under both subject-dependent evaluation with repeated random splits and subject-independent leave one subject out(LOSO) cross-validation show that DRSN architecture consistently outperforms the RefineNet-based baseline, and the triple-channel configuration achieves the best performance (98.88% accuracy). Overall, the variable-channel design enables flexible feature selection to meet diverse application requirements. Full article

Pollen identification is critical for melissopalynology (honey authentication), ecological monitoring, and allergen tracking, yet manual microscopic analysis remains labor-intensive, subjective, and error-prone when multiple grains overlap in realistic samples. Existing automated approaches often fail to address multi-grain scenarios or lack systematic comparison across classical and deep learning paradigms, limiting their practical deployment. This study proposes a subinstance-based classification framework combining YOLOv12n object detection for grain isolation, independent classification via classical machine learning (ML), convolutional neural networks (CNNs), or Vision Transformers (ViTs), and majority voting aggregation. Five classical classifiers with systematic feature selection, three CNN architectures (ResNet50, EfficientNet-B0, ConvNeXt-Tiny), and three ViT variants (ViT-B/16, ViT-B/32, ViT-L/16) are evaluated on four datasets (full images vs. isolated grains; raw vs. CLAHE-preprocessed) for four berry pollen species (Ribes nigrum, Ribes uva-crispa, Lonicera caerulea, and Amelanchier alnifolia). Stratified image-level splits ensure no data leakage, and explainable AI techniques (SHAP, Grad-CAM++, and gradient saliency) validate biological interpretability across all paradigms. Results demonstrate that grain isolation substantially improves classical ML performance (F1 from 0.83 to 0.91 on full images to 0.96–0.99 on isolated grains, +8–13 percentage points), while deep learning excels on both levels (CNNs: F1 = 1.000 on full images with CLAHE; ViTs: F1 = 0.99). At the instance level, all paradigms converge to near-perfect discrimination (F1 ≥ 0.96), indicating sufficient capture of morphological information. Majority voting aggregation provides +3–5% gains for classical methods but only +0.3–4.8% for deep models already near saturation. Explainable AI analysis confirms that models rely on biologically meaningful cues: blue channel moments and texture features for classical ML (SHAP), grain boundaries and exine ornamentation for CNNs (Grad-CAM++), and distributed attention across grain structures for ViTs (gradient saliency). Qualitative validation on 211 mixed-pollen images confirms robust generalization to realistic multi-species samples. The proposed framework (YOLOv12n + SVC/ResNet50 + majority voting) is practical for deployment in honey authentication, ecological surveys, and fine-grained biological image analysis. Full article

Background: Emotion recognition using electroencephalography (EEG) offers a non-invasive means of measuring brain responses to affective stimuli. However, since EEG signals can vary significantly between subjects, developing a deep learning model capable of accurately predicting emotions is challenging. Methods: To address that challenge, this study proposes a deep learning approach that fuses EEG features with demographic information, specifically age, sex, and nationality, using an attention-based mechanism that learns to weigh each modality during classification. The method was evaluated using three benchmark datasets: SEED, SEED-FRA, and SEED-GER, which include EEG recordings of 31 subjects of different demographic backgrounds. Results: We compared a baseline model trained solely on the EEG-derived features against an extended model that fused the subjects’ EEG and demographic information. Including demographic information improved the performance, achieving 80.2%, 80.5%, and 88.8% for negative, neutral, and positive classes. The attention weights also revealed different contributions of EEG and demographic inputs, suggesting that the model learns to adapt based on subjects’ demographic information. Conclusions: These findings support integrating demographic data to enhance the performance and fairness of subject-independent EEG-based emotion recognition models. Full article

Major depressive disorder (MDD) is a prevalent mental health condition that requires accurate and objective diagnostic tools. Electroencephalogram (EEG) signals provide valuable insights into brain activity and have been widely studied for mental disorder classification. In this study, we propose a novel DML + ANFIS framework that integrates deep metric learning (DML) with an adaptive neuro-fuzzy inference system (ANFIS) for the automated diagnosis of major depressive disorder (MDD) using EEG time series signals. Time–frequency features are first extracted from raw EEG signals using the short-time Fourier transform (STFT) and the continuous wavelet transform (CWT). These features are then embedded into a low-dimensional space using a DML approach, which enhances the inter-class separability between MDD and healthy control (HC) groups in the feature space. The resulting time–frequency feature embeddings are finally classified using an ANFIS, which integrates fuzzy logic-based nonlinear inference with deep metric learning. The proposed DML + ANFIS framework was evaluated on a publicly available EEG dataset comprising MDD patients and healthy control (HC) subjects. Under subject-dependent evaluation, the STFT-based DML + ANFIS and CWT-based models achieved an accuracy of 92.07% and 98.41% and an AUC of 97.28% and 99.50%, respectively. Additional experiments using subject-independent cross-validation demonstrated reduced but consistent performance trends, thus indicating the framework’s ability to generalize to unseen subjects. Comparative experiments showed that the proposed approach generally outperformed conventional deep learning models, including Bi-LSTM, 2D CNN, and DML + NN, under identical experimental conditions. Notably, the DML module compressed 1280-dimensional EEG features into a 10-dimensional embedding, thus achieving substantial dimensionality reduction while preserving discriminative information. These results suggest that the proposed DML + ANFIS framework provides an effective balance between classification performance, generalization capability, and computational efficiency for EEG-based MDD diagnosis. Full article

Background/Objectives: Alzheimer’s disease (AD) requires accessible and non-invasive biomarkers that can support early detection, especially in settings lacking specialized expertise. Sonification techniques may offer an alternative way to convey neurophysiological information through auditory perception. This study aimed to evaluate whether human listeners without EEG training can discriminate between sonified electroencephalographic (EEG) patterns from patients with AD and healthy controls. Methods: EEG recordings from 65 subjects (36 with Alzheimer’s, 29 controls) from the Open-Neuro ds004504 dataset were used. Data were processed through sliding-window spectral analysis, extracting relative band powers across five frequency bands (delta: 1–4 Hz, theta: 4–8 Hz, alpha: 8–13 Hz, beta: 13–30 Hz, gamma: 30–45 Hz) and spectral entropy, aggregated across 10 topographic regions. Extracted features were sonified via parameter mapping to independent synthesis sources per frequency band, implemented in an interactive web interface (Tone.js v14.8.49) enabling auditory evaluation. Eight evaluators without EEG experience blindly classified subjects into two groups based solely on listening to the sonifications. Results: Listeners achieved a mean classification accuracy of 76.12% (SD = 17.95%; range: 49.25–97.01%), exceeding chance performance (p = 0.001, permutation test). Accuracy variability across evaluators suggests that certain auditory cues derived from the sonified features were consistently perceived. Conclusions: Parametric EEG sonification preserves discriminative neurophysiological information that can be perceived through auditory evaluation, enabling above-chance differentiation between Alzheimer’s patients and healthy controls without technical expertise. This proof-of-concept study supports sonification as a complementary, accessible method for examining brain patterns in neurodegenerative diseases and highlight its potential contribution to the development of accessible diagnostic tools. Full article

Supervised ECG-based sleep apnea detection typically depends on large and fully annotated datasets, yet the rarity and cost of labeling apneic events often lead to substantial annotation scarcity in practice. This study provides a controlled evaluation of how such scarcity degrades classification performance and, as a key contribution, investigates whether a constrained, morphology-preserving ECG augmentation framework can compensate for reduced apnea-label availability. Using the PhysioNet Apnea–ECG dataset, we simulated seven levels of label retention ($r=5$–$100\%$) and trained a lightweight CNN–BiLSTM model under both subject-dependent (SD) and subject-independent (SI) five-fold protocols. Offline augmentation was applied only to apnea segments and consisted of simple, physiologically motivated time-domain perturbations designed to retain realistic cardiac and respiratory dynamics. Across both evaluation settings, augmentation substantially mitigated performance loss in the low- and mid-scarcity regimes. Under SI evaluation, the mean F1-score improved from 0.57 to 0.72 at $r=5\%$ and from 0.63 to 0.76 at $r=10\%$, with scores at $r=10$–$40\%$ (0.75–0.77) approaching the full-label baseline of 0.79. Temporal and spectral analyses confirmed preservation of P–QRS–T morphology and respiratory modulation without distortion. These results demonstrate that simple and interpretable ECG augmentations provide an effective and reproducible baseline for data-efficient apnea screening and offer a practical path toward scalable annotation and robust single-lead deployment under label scarcity. Full article

Quantitative vascular analysis involves the measurements of arterial tortuosity and branch angle in a region of interest in cerebral arteries to assess vascular risks associated with cerebral aneurysm. The measurements themselves are not a simple process since they are made on the three-dimensional (3D) structures of the arteries. The aim of this study was to develop a deep convolutional neural network (CNN) model to predict a probability score of aneurysm without direct measurements of the artery’s geometry. A total of 204 subjects’ image data were considered. In all, 585 gray-scale three-dimensional (3D) patches with the bifurcations near the center of the patches were extracted and labeled as either an aneurysm or a non-aneurysm class. Three-dimensional CNN architectures were developed and validated for the binary classification of the 3D patches. Accuracy, precision, recall, F1-score, receiver operating characteristics area under the curve (ROC-AUC), and precision recall AUC (PR-AUC) were calculated for test data. Deep learning predictions were compared with vessel geometry measurements. Deep learning probability scores were dichotomized into high-score and low-score groups. For both groups, bifurcation angles and sum-of-angles-metric (SOAM) were calculated and compared. ResNetV2_18 with translation as data augmentation achieved the highest mean ROC-AUC (0.735) and PR-AUC (0.472). The independent t-test indicated that for the bifurcation angle sum feature there was a statistically significant difference (t = −2.280, p-value < 0.05) between the low-score and the high-score groups. In conclusion, we have demonstrated a deep learning-based approach to the prediction of aneurysmal risks in the bifurcation regions of interest. Deep learning predictions were associated with vessel geometry measurements. This suggests that deep learning on 3D patches centered around the bifurcations has the potential to screen bifurcations with a high aneurysm risk. Full article

Background: Periodontitis is a common inflammatory disease characterized by progressive loss of alveolar bone. Cone-beam computed tomography (CBCT) can visualize 3D periodontal bone defects, but its interpretation is time-consuming and examiner-dependent. Deep learning may support standardized CBCT assessment if performance and biological relevance are adequately characterized. Methods: We used the publicly available MMDental dataset (403 CBCT volumes from 403 patients) to train a 3D ResNet-18 classifier for binary discrimination between periodontitis and healthy status based on volumetric CBCT scans. Volumes were split by subject into training (n = 282), validation (n = 60), and test (n = 61) sets. Model performance was evaluated using area under the receiver operating characteristic curve (AUROC), area under the precision–recall curve (AUPRC), and calibration metrics with 95% bootstrap confidence intervals. Grad-CAM saliency maps were used to visualize the anatomical regions driving predictions. To explore phenotype-level biological concordance, we analyzed an independent gingival transcriptomic cohort (GSE10334, n ≈ 220 arrays after quality control) and an independent oral microbiome cohort based on 16S rRNA amplicon sequencing, using unsupervised clustering, differential expression/abundance testing, and pathway-level summaries. Results: On the held-out CBCT test set, the model achieved an AUROC of 0.729 (95% CI: 0.599–0.850) and an AUPRC of 0.551 (95% CI: 0.404–0.727). At a high-sensitivity operating point (sensitivity 0.95), specificity was 0.48, yielding an overall accuracy of 0.62. Grad-CAM maps consistently highlighted the alveolar crest and furcation regions in periodontitis cases, in line with expected patterns of bone loss. In the transcriptomic cohort, inferred periodontitis samples showed up-regulation of inflammatory and osteoclast-differentiation pathways and down-regulation of extracellular-matrix and mitochondrial programs. In the microbiome cohort, disease-associated samples displayed a dysbiotic shift with enrichment of classic periodontal pathogens and depletion of health-associated commensals. These omics patterns are consistent with an inflammatory–osteolytic phenotype that conceptually aligns with the CBCT-defined disease class. Conclusions: This study presents a proof-of-concept 3D deep learning model for CBCT-based periodontal disease classification that achieves moderate discriminative performance and anatomically plausible saliency patterns. Independent transcriptomic and microbiome analyses support phenotype-level biological concordance with the imaging-defined disease class, but do not constitute subject-level multimodal validation. Given the modest specificity, single-center imaging source, and inferred labels in the omics cohorts, our findings should be interpreted as exploratory and hypothesis-generating. Larger, multi-center CBCT datasets and prospectively collected paired imaging–omics cohorts are needed before clinical implementation can be considered. Full article