Search Results (266)

High-dynamic amusement ride conditions involving impacts, rapid rotations, and abrupt posture changes introduce severe motion artifacts that degrade vital sign quality and destabilize physiological state recognition. This study aims to develop an engineering-ready closed-loop framework for robust passenger monitoring and intelligent diagnosis. A multimodal sensing and modeling pipeline was designed to jointly leverage physiological signals such as heart rate and SpO₂ and kinematic measurements, including acceleration, angular rate, velocity, and attitude. Inertial and PPG signals were preprocessed into supervised samples through wavelet multiresolution denoising and coordinate frame unification, while a strapdown inertial navigation system was used to propagate a 12-channel physical quantity sequence. To ensure interpretability and standards compliance, constraints from GB 8408-2018 were translated into executable threshold rules, enabling standards-driven auto-labeling and rule-based early warning. Building on this foundation, three learning modules were developed: a fusion model for high-dynamic heart rate estimation, a CNN–LSTM dynamic-threshold-enhanced network TAPNet for rapid kinematic anomaly screening, and an attention-augmented hybrid model HS-BANet integrating one-dimensional residual blocks, bidirectional LSTM, and multi-head attention for fine-grained arrhythmia classification. Experimental results demonstrated accurate and consistent heart rate estimation with RMSE of 1.18 bpm on HSSH-I and 1.24 bpm on the independent HSSH-II set, strong agreement with training and testing correlations of 0.9928 and 0.9865, and near-zero bias in Bland–Altman analysis. TAPNet achieved 96.9% validation accuracy and 98.2% test accuracy for kinematic anomaly recognition, maintaining robust generalization under class imbalance. HS-BANet enabled multi-class identification of PVC, PAC, VT, SVT, and AF, achieving an accuracy of 92.37%, an F1-score of 86.87%, a precision of 88.45%, a sensitivity of 88.14%, and a specificity of 89.42%. Overall, the proposed two-stage multimodal closed-loop—fast, interpretable early warning based on physical quantity thresholds followed by fine-grained diagnosis from physiological signals—supports stable feature extraction and reliable decision-making under strong motion artifacts and non-stationary dynamics, balancing responsiveness and diagnostic credibility, while showing potential for practical safety early warning and future deployment-oriented operational support in amusement ride scenarios. Full article

Emotion recognition plays a crucial role in human–computer interaction, health monitoring, and affective computing by analysing physiological signals. Despite recent advancements, current research still faces challenges, including the lack of effective fusion strategies for diverse physiological modalities, difficulties in handling high-dimensional feature representations, and limited use of efficient temporal modelling techniques to capture complex emotional patterns. This study proposes a deep learning-based approach that fuses multiple physiological modalities, including Electroencephalography (EEG), Electrooculography (EOG), Electromyography (EMG), Galvanic Skin Response (GSR), Respiratory Rate (RR), Skin Temperature (SKT), and Photoplethysmography (PPG), to improve emotion recognition. Arousal and valence ratings were binarized into two classes (low/high) using a threshold of 4.5, formulating a binary classification problem. In addition to utilising Bidirectional Long Short-Term Memory (Bi-LSTM), the study employs Temporal Convolutional Networks (TCN), a widely used approach for time-series analysis, to efficiently capture temporal dependencies. The proposed model optimises feature selection through channel-wise strategies, incorporates advanced learning rate scheduling, and reduces computational overhead. Furthermore, window-wise, block-wise, and trial-wise evaluation protocols were investigated to assess the impact of temporal information leakage on emotion recognition performance. Using the DEAP dataset for validation, the proposed TCN-based approach achieved classification accuracies of 88.42% for valence and 86.35% for arousal under an overlapping block-wise evaluation protocol, demonstrating improved performance in binary emotion recognition and highlighting the importance of leakage-aware model assessment. Full article

To address the issues of low perception accuracy and poor robustness in traditional motion recognition methods within complex walking environments for visually impaired individuals, this study utilizes multi-modal data, including ECG, PPG, and IMU, for classification. Regarding the low filtering efficiency of multi-modal data, an improved wavelet filtering algorithm based on LSTM is proposed. To further enhance classification accuracy, this paper introduces a motion recognition method for the blindfolded mobility simulation based on an Attention-based Two-Stream Deep Fusion Convolutional Neural Network (ATS-DFCNN). The proposed method constructs a two-stream heterogeneous feature extraction architecture by synchronously collecting tri-axial motion signals and physiological signals from subjects. A 1D-CNN is employed to capture the spatial geometric features of limb movements, while a hybrid CNN-GRU network is utilized to mine the temporal evolution patterns of physiological stress. Furthermore, an attention mechanism is introduced to achieve dynamic weighted fusion at the feature level, which strengthens critical motion features and suppresses environmental noise. Experiments were conducted with 10 subjects simulating the movements of visually impaired individuals, covering typical actions such as walking, standing, climbing stairs, descending stairs, and falling. The results demonstrate that the proposed adaptive filtering algorithm achieves an AUC of 0.942, significantly improving feature distinctiveness compared to traditional algorithms. The ATS-DFCNN model achieved an average recognition accuracy of 92.2% across five activity categories, representing a 4.8% performance increase over single IMU modal classification. Particularly in fall detection, the model effectively reduces false alarms through physiological feedback and accurately infers motion intentions, providing reliable technical support for the safety monitoring of intelligent walking-aid systems. Full article

Objective: Rapid vertical manoeuvres and intermittent vibration in autonomous electric vertical take-off and landing (eVTOL) aircraft can provoke pronounced psychological anxiety in passengers. To address this, we propose a closed-loop adaptive system that integrates an in-ear wearable sensor with dynamic regulation of the cabin microenvironment, enabling real-time monitoring of each passenger’s autonomic state and delivering individualised mitigation through a continuous sense–analyse–intervene–feedback loop. Methods: The system is built around a pair of custom in-ear modules that integrate dual-wavelength photoplethysmography (PPG; 525 nm green and 940 nm infrared), galvanic skin response (GSR), and a six-axis inertial measurement unit (IMU) sampled at 200 Hz. To suppress the 20–80 Hz vibration generated by the distributed electric propulsion system, a compliant silicone damping sleeve attenuates high-frequency components at the hardware level, while a Kalman filter fuses the IMU and PPG streams and an adaptive notch filter removes residual rotor harmonics. The pipeline raises the heart-rate-variability (HRV) signal-to-noise ratio (SNR) to 24.1 dB, with a Pearson correlation of 0.96 against a medical-grade chest strap. A hybrid CNN–LSTM network—two convolutional layers (32 filters each) followed by two LSTM layers (128 hidden units)—predicts impending anxiety from HRV time-domain features (RMSSD, pNN50) and frequency-domain features (LF/HF ratio), triggering intervention 8.2 s in advance on average. According to the predicted anxiety level (mild/moderate/severe), a fuzzy controller modulates transcutaneous auricular vagus nerve stimulation (1–5 mA), the binaural-beat frequency (4–8 Hz, theta band), and the cabin lighting colour temperature (2700–6500 K) in real time. The intervention parameters are continuously refined by SPSA-based stochastic optimisation of the HRV recovery rate (step size 0.01; updated every 30 s). Results: In a randomised controlled experiment conducted in a simulated flight environment (N = 50; aged 22–45 years; 1:1 sex ratio), the active group reached physiological recovery in 52.3 s on average, compared with 98.6 s for the sham-controlled group—a 47% reduction (Cohen’s d = 1.24, p < 0.001). User acceptance reached 94%. Conclusions: The proposed in-ear platform enables closed-loop adaptive regulation of anxiety in the eVTOL cabin and overcomes the limitations of conventional passive mitigation strategies. By combining vibration-tolerant physiological sensing with multimodal environmental control, the work offers a practical pathway for improving passenger experience in urban air mobility and provides a useful reference for human-factors standards governing autonomous aircraft. Full article

Non-contact physiological monitoring based on remote PPG (rPPG) offers a viable alternative for the care of pediatric populations, particularly for children with cerebral palsy (CP) who present unique communication and mobility challenges. This paper presents a review of the literature on the use of rPPG for the estimation of vital signs and its application in emotional monitoring. Following the PRISMA 2020 guidelines as a methodological framework for searching and filtering, an exhaustive search was conducted in the IEEE Xplore and Scopus databases covering the period from 2017 to 2024. A total of 35 studies were selected for analysis. The review examines the evolution of rPPG algorithms—from classical mathematical approaches to recent deep-learning-based architectures—identifying critical technical challenges such as motion artifacts caused by spasticity and variations in lighting conditions. The results reveal that while rPPG has reached technical maturity for monitoring core physiological parameters such as heart rate, its application to robust emotion detection in children with CP remains limited. The main limitation identified across the surveyed literature is the critical scarcity of public or clinical datasets featuring pediatric CP cohorts. Finally, the potential of multimodal integration—combining rPPG with eye-tracking and wearable sensors—is discussed as a promising pathway toward objective emotional monitoring. Such an approach could enhance communication, support rehabilitation processes, and ultimately improve the quality of life of children with cerebral palsy and their caregivers. Full article

This paper proposes a two-stage hybrid framework for biosignal quality validation that produces beat-level or segment-level labels for real-time filtering and offline dataset curation. The framework is quantitatively validated exclusively on ECG data. Its modular architecture is designed to extend to further non-stationary periodic biomedical time-series signals including photoplethysmography (PPG), impedance cardiography (ICG), phonocardiography (PCG), electromyography (EMG), and electroencephalography (EEG) through modality-specific parameter adaptation; however, this broader applicability currently reflects architectural extensibility rather than experimentally validated performance. A prerequisite is synchronized acquisition of the primary biosignal together with inertial motion sensing (IMU/accelerometer) and electrode impedance or lead-off status, with the IMU positioned near the sensing electrodes. The first stage performs sensor-integrity gating to reject intervals corrupted by motion or poor electrode contact. The second stage applies software signal quality indices to the remaining beats, including physiological plausibility constraints (R to R peaks analysis), DTW-based morphological consistency against adaptive templates, frequency domain SNR estimation, and baseline wander quantification. This study systematically evaluates and compares the classification performance of six complementary sensor-level and software-based signal quality assessment methods. When integrated within the proposed hybrid framework, validation against expert-annotated ECG quality labels from 20 healthy participants demonstrates high methodological classification accuracy (98.1%), achieving approximately a 98% F1-score, 99% sensitivity, and 97% specificity. Prospective validation on patient populations with cardiovascular pathology is identified as a necessary step toward clinical deployment. This modular approach improves the reliability of downstream analysis by preventing corrupted data from entering feature extraction and model training pipelines, enabling more stable physiological monitoring in free-living conditions, reducing false alarms in continuous monitoring applications, and generating higher-quality datasets for AI-based diagnostic systems. Full article

Background: Post-surgical cardiovascular monitoring places a heavy information burden on clinical teams, requiring the rapid synthesis of patient history, intraoperative data, monitoring streams, and surgical outcome evidence. Existing clinical decision support systems handle this integration poorly, and most offer little visibility into their reasoning. We present a Retrieval-Augmented Generation (RAG) architecture designed specifically for this domain, with a focus on evidence traceability and practical workflow integration. Methods: We describe a three-layer RAG architecture comprising a retrieval layer that creates 768-dimensional representations of clinical scenarios; an augmentation layer using a stacking ensemble (Random Forest and XGBoost base learners with a logistic-regression meta-learner) to integrate patient-specific data with retrieved evidence and produce calibrated probability estimates; and a generative layer using a fine-tuned BERT classifier together with Gemini 2.5 Pro to synthesise actionable clinical recommendations. Components were prototyped on publicly available, de-identified data from MIMIC-III and the MIMIC-III-Ext-PPG benchmark to verify pipeline integrity. Proposed Evaluation Framework: This paper presents a system architecture rather than a clinically validated implementation. We outline a structured evaluation framework to assess the technical performance and clinical applicability of the RAG architecture, encompassing the technical validation of system components, expert assessment of clinical workflow integration potential, and analysis of interpretability features essential for healthcare deployment. Specific technical targets include retrieval precision >90% for relevant evidence, query response time <3 s, and a clinical appropriateness rating of >85% from expert review. Conclusions: We describe a RAG architecture for post-surgical cardiovascular monitoring in which every recommendation is linked to retrievable source documents, making the reasoning visible and challengeable. A structured evaluation framework is proposed to guide the system towards clinical validation. Full article

Arrhythmia is a common and potentially life-threatening cardiovascular condition. Photoplethysmography (PPG) has emerged as a noninvasive alternative to electrocardiography for cardiac rhythm monitoring, yet most PPG-based methods remain limited to binary classification. In this study, a new deep learning approach is suggested for categorizing six arrhythmia types from PPG data: sinus rhythm (SR), premature ventricular contraction (PVC), premature atrial contraction (PAC), ventricular tachycardia (VT), supraventricular tachycardia (SVT), and atrial fibrillation (AF). The raw PPG signal is enhanced by extracting its first and second derivatives to capture morphological features not readily apparent in the original signal. A hybrid architecture, VGG-BiLSTM, is utilized, merging VGG convolutional layers for spatial features extraction with bidirectional long short-term memory layers for modeling temporal dependencies. A stratified data splitting strategy is further adopted to address class imbalance across arrhythmia types. A publicly available dataset containing 46,827 PPG segments from 91 individuals was employed to assess the effectiveness of the suggested technique. The method yielded an overall accuracy, sensitivity, specificity and F1 score of 88.7%, 78.5%, 97.6% and 80.5% correspondingly. Full article

Chronic stress is a time-dependent condition characterized by sustained dysregulation across neural, autonomic, and endocrine systems, with important consequences for both health and socioeconomic outcomes. Unlike acute stress, which is typically characterized by short-lived physiological activation, chronic stress reflects an accumulated allostatic load and a longer-term recalibration of stress response systems. Recent advances in physiological sensing and artificial intelligence (AI) have supported the development of computational approaches for chronic stress detection using electroencephalography (EEG), heart rate variability (HRV), photoplethysmography (PPG), electrodermal activity (EDA), and wearable multimodal platforms. This narrative review examines current AI-based studies through three main inferential paradigms: resting baseline dysregulation, longitudinal physiological monitoring, and reactivity-based inference. Across modalities, classical machine learning (ML) methods, particularly support vector machines (SVMs) and tree-based ensembles, remain the most commonly used approaches, largely because available datasets are small and most pipelines still depend on engineered features. Deep learning (DL) methods are beginning to emerge, but their use remains constrained by the lack of large, standardized, longitudinal datasets specifically designed for chronic stress research. Major challenges include ambiguity in stress labeling, limited longitudinal validation, circadian confounding, inter-individual variability, and small cohort sizes. Future progress will depend on standardized datasets, biologically grounded multimodal integration, hybrid baseline-reactivity modeling, adaptive personalization, and more interpretable AI systems. Greater emphasis is also needed on clinical relevance and generalizability if AI-based chronic stress monitoring is to move beyond experimental settings. Full article

Remote photoplethysmography (rPPG) extracts physiological signals from facial videos by analyzing subtle skin color variations caused by blood flow. While this technology enables contactless health monitoring, it also raises privacy concerns because facial videos reveal both identity and sensitive biometric information. Existing privacy-preserving techniques, such as blurring or pixelation, degrade visual quality and are unsuitable for practical rPPG applications. This paper presents rePPG, a framework that inserts a desired rPPG signal into facial videos while preserving the original facial appearance. The proposed method disentangles facial appearance and physiological features, enabling replacement of the physiological signal without altering facial identity or visual quality. Skin segmentation restricts modifications to skin regions, and a cycle-consistency mechanism ensures that the injected rPPG signal can be reliably recovered from the generated video. Importantly, the extracted rPPG signals are evaluated against the injected target physiological signals rather than the subject’s original physiological state, ensuring that the evaluation measures signal rewriting accuracy. Experiments on the PURE and UBFC datasets show that rePPG successfully embeds target PPG signals, achieving 1.10 BPM MAE and 95.00% PTE6 on PURE while preserving visual quality (PSNR 24.61 dB, SSIM 0.638). Heart rate metrics are computed using a 5-second temporal window to ensure a consistent evaluation protocol. Full article

Hypertension is a leading global health risk and requires accurate and continuous monitoring for effective management. Although photoplethysmography (PPG) is a promising non-invasive modality for cuffless blood pressure (BP) assessment, many existing approaches (especially raw-signal deep learning) are vulnerable to data leakage, overfitting on small datasets, limited interpretability, and poor performance on minority BP stages. To address these limitations, we propose a robust and physiologically grounded framework for multi-class BP stage classification based on interpretable PPG features. Our approach centers on a comprehensive multi-domain feature engineering pipeline that extracts 124 PPG features, including demographic, morphological, functional decomposition, spectral, nonlinear dynamics, and clinical composite indices. We apply rigorous preprocessing and feature selection prior to model training. We validate the framework on two datasets: PPG-BP dataset (657 segments, 4 classes) for benchmarking and PulseDB (283,773 segments, 3 classes) to assess scalability. We evaluate the proposed framework using a segment-level train/test split, appropriate for assessing intra-subject BP tracking after initial personalization. For the PulseDB dataset, this follows the protocol established by the dataset creators, while for the PPG-BP dataset, it enables direct comparison with prior work given practical dataset constraints. On PPG-BP, LightGBM trained on the selected features achieved macro-F1 = 0.78 and accuracy = 0.74, outperforming comparable deep-learning models. On the PulseDB, a custom Residual MLP achieved accuracy = 0.81 and macro-F1 = 0.79, supporting generalization at scale. These results show that the proposed feature-based approach can outperform complex end-to-end deep-learning models on small datasets while providing improved interpretability. This work establishes a reliable and transparent pathway toward clinically viable continuous BP staging, moving beyond black-box models toward physiologically grounded decision support. Ablation analysis reveals that engineered features provide most of the predictive power (F1 = 0.911), while raw PPG features alone achieve modest performance (F1 = 0.384). For the minority hypertension stage 2 (HT-2) class, a bootstrap 95% confidence interval of [0.762, 1.000] is reported, reflecting uncertainty due to limited sample size. Full article

Continuous non-invasive blood pressure monitoring holds significant promise for cardiovascular disease management, yet cuff-based methods remain limited by their intermittent nature. Machine learning approaches leveraging photoplethysmography (PPG) and electrocardiography (ECG) signals present compelling alternatives, though questions persist about which signal type contributes more predictive value. This study compares traditional machine learning models, ensemble methods, and deep learning architectures for estimating systolic blood pressure from physiological waveforms. We extracted 55 features from PPG and ECG recordings of 100 subjects in the MIMIC-III Waveform Database, yielding 3000 segments with invasive arterial blood pressure as ground truth. Data splitting was performed at the subject level (70/15/15 train/validation/test) to prevent data leakage. Evaluation included regression metrics, British Hypertension Society grading, SHAP-based explainability, and ablation studies. Among all models, LightGBM achieved the best performance with mean absolute error of 15.97 mmHg, placing it at BHS Grade D. While SHAP analysis showed ECG features contributing 54.7% of importance versus 45.3% for PPG, our ablation study revealed that PPG-only models achieved comparable performance (MAE 15.97 vs. 16.23 mmHg), with the difference not statistically significant (p = 0.226). These results suggest that PPG-only wearable devices are viable for blood pressure estimation, as adding ECG features provides no statistically significant improvement. However, all configurations achieved only BHS Grade D, indicating that personalized calibration may be necessary for clinical acceptability. Full article

Respiratory rate (RR) is a critical vital sign for the early detection of hypoxia and respiratory deterioration, yet its continuous monitoring remains challenging in clinical environments. Photoplethysmography (PPG) provides a non-invasive source of physiological information from which respiratory dynamics can be inferred. In this study, numerical physiological features derived from PPG data were used to comparatively evaluate multiple deep learning models for respiratory rate estimation. Fixed-length sliding windows were constructed from the dataset and used to train five neural network architectures: a Deep Feedforward Neural Network (DFNN), unidirectional and bidirectional Recurrent Neural Networks (RNN, Bi-RNN), and unidirectional and bidirectional Long Short-Term Memory networks (LSTM, Bi-LSTM). Model performance was assessed using mean absolute error (MAE), root mean squared error (RMSE), coefficient of determination (R²), and computational runtime. Results indicate that models incorporating temporal dependencies outperform the static feedforward baseline, achieving MAE values as low as 0.521 breaths/min, making them competitive with or lower than previously reported PPG-based approaches. These findings highlight the effectiveness of temporal deep learning models for respiratory rate estimation from PPG-derived numerical features and provide insight into accuracy–efficiency trade-offs relevant to real-time monitoring applications. Full article

In autonomous driving, motion sickness (MS) arises from physical or visual stimuli, or a combination of both. However, objective quantification of MS level (MSL) remains limited beyond questionnaire-based assessments. Using multimodal human signals (physiological and behavioral) collected in an autonomous driving simulator, this study addresses the association between these signals and MSL, across these MS types, by (i) screening and curating a decade of human-signal MS studies (HS-Set) to establish a data-driven foundation for selecting target sensor domains and features, (ii) constructing a dataset with subjective measures of MSL (fast motion sickness scale and simulator sickness questionnaire (SSQ)), alongside human signals (electroencephalogram (EEG), photoplethysmogram (PPG), electrodermal activity (EDA), skin temperature, and head/eye movement), (iii) conducting a correlation analysis between MSL and the identified features from HS-Set, and (iv) quantifying multivariable contributions at the feature and sensor domains through an explainable boosting machine (EBM). Key correlations include head amplitude/energy (pitch/surge) with SSQ total/oculomotor, eye entropy with nausea/oculomotor (positive), and EDA with nausea (negative). The EBM-based contribution analysis highlights EEG connectivity and head kinematics as dominant contributors; excluding EEG, the interpretability of single-domain models remains limited. Additionally, a combination of Head, PPG, and EDA domains retains over 80% of the full model’s interpretability. Full article

Remote photoplethysmography (rPPG) enables noncontact estimation of vital signs, particularly heart rate, by analyzing subtle periodic skin color variations in facial videos. While deep learning has advanced rPPG signal extraction, existing methods rely on carefully designed task-specific architectures that are costly to develop and generalize poorly. In this work, we demonstrate that the general video foundation model VideoMAE v2 can be effectively adapted to the rPPG signal regression task by introducing only a lightweight adapter, without modifying its pretrained backbone. We freeze the entire VideoMAE v2 encoder and introduce a Temporal Difference Convolutional Adapter to capture the subtle interframe intensity differences. To address the mismatch between VideoMAE v2′s short input window (16 frames) and the long temporal context typically required for robust rPPG extraction (e.g., 160 frames), we adopt an overlapping sliding window strategy for segmented inference and reconstruct the full signal through weighted temporal aggregation. On the COHFACE and UBFC-rPPG datasets, our method achieves mean absolute errors (MAEs) of 0.90 and 1.55, reducing the error by more than 55% and 42%, respectively, compared to PhysFormer (2.00 and 2.70). Furthermore, on challenging real-world datasets such as BUAA-MIHR, which features strong illumination variations, and VIPL-HR, which involves significant head movements, our approach achieves MAEs of 6.68 and 8.23, respectively, despite incorporating no task-specific robustness modules. These results demonstrate stable rPPG signal recovery and validate the feasibility of leveraging general video foundation models for physiological signal perception. Full article