IEEE Transactions on Biomedical Engineering

Millimeter-wave radar can quietly monitor health and behavior at home, which is vital for supporting people living with dementia. Most studies, however, remain limited to short-term testing in controlled spaces. Real-world deployment requires robust activity classification as a prerequisite: vital-sign and behavioral sensing require fundamentally different processing pipelines, and absent periods need to be reliably distinguished from stationary states. Bridging the critical gap between controlled laboratory demonstrations and continuous home monitoring, this paper introduces a self-adapting radar framework that extracts meaningful behavioral segments from massive, unconstrained real-world data. The system performs continuous real-time activity classification (stationary, walking, and absent) and target localization, selectively directing downstream processing to the most informative segments. It addresses key real-world deployment challenges including adaptive thresholding across subjects and environments, and walking detection under naturalistic activity conditions. Prior to integration with the Minder platform, the system was validated in a fully instrumented studio apartment against ground truth. Across 12 subjects, the system achieved an overall classification accuracy of 0.98, with F1 scores of 0.99 for absence and stationary states, and 0.95 for walking. Event-based evaluation yielded a per-subject walking sensitivity of 0.916{+/-} 0.058 and F1 score of 0.935 {+/-}0.030. Localization root mean square error during movement was 0.40 m. The results demonstrate reliable performance suitable for transitioning to long-term real-world home deployment.

This paper presents experimental evidence that alpha-band EEG signals can be reliably detected from an in-ear electrode during physical activity, enabling fatigue monitoring in dynamic, real-world conditions such as sports. We collected an EEG dataset using a custom-designed, compact wearable system measuring only 20 mm in diameter, integrated inside the earphone. It supports five channels, four head electrodes (T3, C3, C4, T4) and one in-ear electrode, allowing simultaneous multi-site recordings. Recordings were made while a participant engaged in a controlled cycling protocol designed to induce physical fatigue. We demonstrated a direct relationship between alpha power and entropy in EEG data recorded from both the head and ear, during both activity and rest. To our knowledge, this is the first study to demonstrate in-ear alpha power tracking during active physical movement for sports-related fatigue monitoring. These findings open new possibilities for compact, wearable EEG systems in athletic and high-performance settings, where traditional EEG setups are impractical

Recent advancements in wearable sensors and machine learning show promise for estimating lower-body joint torques outside of laboratory settings. Inertial Measurement Units combined with Convolutional Neural Networks have proven effective for this task. However, the impact of different input data types and formats remains underexplored. This study investigates how variations in input data influence the prediction of lower-body joint torques during walking. Results indicate that while dataset choice causes only minor differences in prediction performance, the overall quality of the dataset plays a more critical role than the specific input variables in achieving accurate torque predictions using wearable sensors.

Accurate pose estimation is central to automated infant General Movements Assessment during the fidgety period, when subtle limb movements, particularly at distal joints inform neurodevelopmental risks. Robust 2D pose tracking from handheld videos remains challenging in real-world settings, where occlusion, rapid motions, and visually ambiguous smaller joints frequently compromise anatomical accuracy. We present CRADLE, a clinically motivated, anatomy-aware post-processing pipeline designed to refine infant 2D movement trajectories across 24-anatomocal landmarks detected by our DeepLabCut-trained model. CRADLE integrates segment-length constraints, velocity-based anomaly detection, anatomically constrained interpolation, and Kalman filtering to correct both large localization failures and subtle persistent joint misplacements without relying primarily on confidence scores. Evaluations against conventional Confidence-Thresholding using Mean Absolute Error (MAE), {Delta}MAE, average Percentage of Correct Keypoints, and net keypoint correction rate showed consistently reduced or preserved error while maintaining accurate trajectories, with the strongest gains achieved at clinically important distal joints. Mean improvements reached up to 5 pixels for some smaller distal landmarks, large-magnitude corrections occurred more often than with Confidence-Thresholding, and well-localised joints remained largely unaffected. Positive net correction rates across metacarpophalangeal and metatarsophalangeal distal-landmarks further confirmed a favourable correction-degradation balance. By improving pose trajectory quality, CRADLE enhances the reliability of downstream movement analysis.

Biomechanical simulations of complex hand motions remain scarce, due to challenges that span computation and data acquisition. Using a computer vision-based motion capture approach, a 23-degree of freedom musculoskeletal model, and direct collocation optimization, we performed muscle-driven simulations to track hand kinematics from 7 participants performing American Sign Language gestures. While proximal joints were tracked accurately, interphalangeal joint tracking was significantly worse, with a consistent flexion bias. Modifications to finger extensor muscle paths that incorporated the dual-inserting nature of the extensors improved accuracy, suggesting better representation of extensor force distribution across distal joints may be necessary for accurate hand simulations.

Spatial attention is critical for solving the cocktail party problem, a longstanding problem in neuroscience and artificial speech recognition. The ability to decode where humans are attending in a cocktail party like scene would empower applications in brain computer interfaces and assistive devices such as hearing aids. Here we demonstrate that, in an overt attention task, the attended spatial location can be decoded robustly from single trial hemodynamic responses, using a wearable whole head high density fNIRS system. We also identify critical brain regions that make the highest contribution to decoding accuracy. Specifically, we find that decoding based on a small fraction of channels within the left and right inferior parietal lobule (IPL), achieve maximal decoding accuracy comparable to all channels. These results open the way for the design of novel BCIs and assistive devices integrated with fNIRS, that can be steered by spatial attention.

Various ankle-foot conditions (e.g., fractures, diabetic foot ulcers, and post-surgical recovery) require periods of complete non-weightbearing followed by gradually increasing partial loadings. However, existing assistive devices often provide inconsistent or uncomfortable offloading during gait. Additionally, prolonged proximal leg offloading can contribute to muscle atrophy, reduced bone density, and overuse of other body segments. We present a novel offloading ankle-foot orthosis (OLAFO) designed to overcome these limitations. The OLAFO features a patient-specific load-bearing shank brace, designed through a digital workflow and fabricated from a 3D-printed core reinforced with carbon-fiber composite lamination. Interlocking serrated side struts, adjustable in 2 mm increments, modulate load sharing between the shank and plantar surfaces. Furthermore, the OLAFO incorporates contact plates with a rocker profile informed by roll-over-shape measurements to support forward progression and gait symmetry. Proof-of-concept biomechanical verification in one able-bodied participant evaluated complete offloading, five partial-loading levels, and normal gait using a pressure walkway to compute vertical ground reaction forces and impulses. In complete offloading, the affected foot generated no contact pressures. Across partial-loading levels, the foot impulse increased from 14% to 53% of the total load and scaled linearly with strut height adjustments, supporting clinician-prescribed loading increments. Contralateral stance duration increased only modestly compared to commonly used assistive devices, indicating reduced compensatory loading on the intact limb. These findings demonstrate the proof-of-concept feasibility of the OLAFO, highlighting its potential for verifying full offloading and prescribing partial-loading targets during rehabilitation. Future research will evaluate performance across patient populations and clinical rehabilitation tasks.

Brain-machine interfaces (BMIs) for vision restoration require models that accurately simulate the anatomy and electrical properties of visual pathways. However, current models focus only on isolated structures, such as the retina or brain, and overlook surrounding tissues. Here, we present a comprehensive computational model of the human head, incorporating the entire visual pathway--including the eye, optic nerve, and brain--along with critical neighboring tissues such as the orbit, paranasal sinuses, enabling precise simulations. Validation using human and large animal data demonstrated a strong correlation between the simulated and measured electrical potentials. Component elimination analysis revealed that the optimized comprehensive model outperformed simplified versions. The models utility was demonstrated through multiple applications: (1) comparative analysis of electrical neuromodulation technologies for optic neuropathy, revealing the filed intensity limitations of noninvasive approaches and the safety concerns of invasive intraorbital approach; (2) identification of optimal stimulation site, revealing that transnasal stimulation at the optic chiasm outperformed traditional approaches; and (3) in silico design of electrode arrays for optic nerve prosthetics, demonstrating theoretical advantages in invasiveness and visual field coverage compared to existing retinal and cortical prosthetics. This validated and versatile computational resource supports the development of neuromodulation strategies and visual BMI technologies.

Wearable digital health technologies may complement traditional gait assessments in amyotrophic lateral sclerosis (ALS) by sensitively capturing real-world mobility changes. In this study, we validated six digital gait metrics derived from ankle-worn sensors in a natural history cohort of 182 individuals with ALS. Investigated metrics correspond to various aspects of gait, including volume, speed, intensity, similarity, variability, and fragmentation. Longitudinal analyses showed significant declines in step count, peak cadence, stride intensity, and stride similarity, with increasing stride duration variability and walking fragmentation over 52 weeks. Many participants exhibited greater relative change in the gait metrics than the self-reported ALS Functional Rating Scale-Revised (ALSFRS-RSE). Stratified analyses revealed that digital metrics captured significant functional decline even in participants with stable walking scores on the ALSFRS-RSE. These findings support the potential utility of these metrics for disease monitoring in ALS clinical care and trials.

Aortic stiffness is both heterogenous and anisotropic. Current non-invasive methods to estimate aortic stiffness are limited to characterizing the aortic tissue as isotropic due to the lack the techniques required to extract multi-axial strain from 3D dynamic images. Vascular deformation mapping (VDM) is a nonrigid image registration technique which has thus far been applied to map aortic growth using longitudinal imaging. In this study, we propose to use VDM to assess 3D aortic deformation by mapping diastolic and systolic images. During image registration process, penalty parameters are employed to fine-tune image alignment and penalize non-physiological deformations. These penalty parameters must be calibrated to ensure that VDM successfully reproduces multi-axial aortic motion patterns in health and disease. In this paper, we developed a calibration pipeline for these parameters using synthetic data. A rotation-free shell model was used to generate physics-based synthetic data on aortic motion incorporating patient-specific geometries, root motion, and blood pressure from a cohort of 14 subjects (healthy, Marfans syndrome and thoracic aortic aneurysm). An error metric was defined to quantify the quality of the VDM results. Furthermore, a k-means clustering technique was used to categorize the subjects into three clusters based on ascending aortic motion. Optimal penalty parameters were identified for each of the three clusters. The results indicated that patient clusters with smaller aortic root motion required larger rigidity penalty values. The calibrated parameters successively reduced errors in 3D displacement and multi-axial stretch compared to un-optimized VDM predictions, enhancing the accuracy of capturing aortic deformation from dynamic images. Among the different aortic regions, the ascending thoracic aorta exhibits the largest error reduction.

Quantitative eye movement analysis is important for neuro- logical diagnostics, yet existing video-oculography (VOG) systems typ- ically require calibration, device-specific settings, or accurate gaze la- bels. We present VOGeo-Gaze, a real-time, calibration-free, geometry- aware neural network that estimates gaze by reconstructing anatomi- cally meaningful eyeball parameters from image features. The method combines segmentation-driven projection geometry, a refraction-aware pupil correction module, and temporal anatomical stabilization, so gaze is derived from interpretable eye geometry rather than direct angular regression. Trained only on the public TEyeD dataset with weak gaze supervision, VOGeo-Gaze was evaluated on 116 clinical recordings from 17 patients and 19 healthy subjects using EyeSeeCam, a clinical gold- standard VOG system. It achieved median absolute angular errors of 0.33{whitebullet} horizontally and 0.35{whitebullet} vertically, with nearly 92% of recordings below 1{whitebullet} error while operating at >300 FPS. These results demonstrate sub-degree clinical gaze estimation without subject-specific calibration, camera intrinsics, or accurate gaze labels, providing a scalable and inter- pretable alternative to conventional VOG pipelines. Code is available at https://github.com/DSGZ-MotionLab/VOGeo-Gaze.

PurposeTo address the limitations of single-distance, 1D performance metrics in RF coil design. This work introduces a multi-objective, volume-of-interest (VOI) based analysis to systematically characterize the trade-offs between power efficiency, pSAR efficiency, and homogeneity as a function of dipole length (l) and distance-to-load (d) for multiple dipole geometries and target anatomies. MethodsElectromagnetic simulations of straight and end-meandered dipole antennas were performed with varying lengths (100-500 mm) and distance-to-load (1-81 mm) over three anatomical targets (prostate, kidney, heart). Homogeneity, power efficiency, pSAR efficiency, and load sensitivity performance metrics were calculated within each anatomical VOI. Inter-element coupling at variable d was assessed in a 3-element array, and a subset of single-element simulations was experimentally validated using B1+ mapping. ResultsA fundamental trade-off was found between power efficiency and pSAR efficiency. Optimal power efficiency was achieved with shorter dipoles (150 mm < l < 300 mm) closer to the sample (d < 30 mm), while optimal pSAR efficiency and homogeneity were achieved with longer dipoles at further from the sample (d > 60 mm). Inter-element coupling increased with distance-to-load but could be managed by increasing element spacing. Experimental measurements were in good agreement with simulation trends. ConclusionIncreasing distance-to-load to 40-60 mm, compared with commonly used distances of 20-30 mm, offers a practical strategy for improving pSAR efficiency and homogeneity with a minimal decrease in power efficiency. This work provides a quantitative analysis that enables RF coil designers to make informed, data-driven decisions when developing next-generation body arrays and suggests that unshielded end-meandered dipoles could be an optimal transmit element geometry.

Tissue-level mechanical stimuli are primary drivers of tissue adaptation and can be optimised during conservative treatments to improve treatment outcomes for many highly prevalent musculoskeletal conditions. Current laboratory-based technologies limit our ability to connect conservative interventions such as exercise and movement modification with muscle, joint, and tissue-level mechanics, in natural environments. We introduce a physics-informed neural network (PINN) to estimate clinically relevant biomechanics from smart garments. By accounting for physiological dynamics of neural activation and muscle contraction, the PINN accurately predicted hip joint angles (RMSE <6 degrees), moments (RMSE 0.12 N*m/kg to 0.30 N*m/kg), and joint forces (RMSE 6 to 16%) from three inertial measurement units and four electromyographic sensors. We demonstrated that the trained PINN can be combined with a smart garment to estimate hip biomechanics, in real-time, during a gait retraining intervention aimed at modifying joint loading to treat hip osteoarthritis. The developed PINN and smart garment system may be adapted and generalised for personalised management or rehabilitation of a broad range of musculoskeletal diseases and injuries, in clinical, home, workplace, and sporting environments.

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease affecting more than 450,000 individuals worldwide and is frequently diagnosed more than 12 months after symptom onset, delaying intervention during a critical early window. Because up to 80% of patients develop dysarthria within two years, subtle changes in speech provide a signal of early bulbar motor neuron degeneration. However, existing speech-based systems rely on supervised classification trained on limited datasets, achieving moderate sensitivity and depending heavily on labeled disease examples, which restrict scalability and early detection. This study introduces SPEAK-NORM, the first-ever normative speech modeling framework for early ALS diagnosis, which learns age- and sex-conditioned motor-speech distributions exclusively from healthy individuals. A conditional variational autoencoder models coordination of hypoglossal, laryngeal, and respiratory motor pathways, and deviation from this healthy manifold is quantified through latent representations and reconstruction error to form a 354-dimensional profile. A calibrated linear Support Vector Machine performs subject-level classification under subject-disjoint validation. On the VOC-ALS database (n = 153), SPEAK-NORM achieves 98% accuracy with balanced sensitivity and specificity, significantly outperforming established clinical acoustic indices and prior systems. The framework maintains strong performance under cross-task generalization and when retrained on healthy controls in independent dementia and Parkinson disease cohorts, demonstrating disease-specific deviation patterns rather than generic neurodegenerative change. Spectral, temporal, and latent separations further support interpretability. By modeling healthy speech instead of memorizing disease examples, SPEAK-NORM enables scalable early neuromotor screening using recording devices, with potential to support earlier diagnosis, differential classification, and monitoring of ALS progression.

BackgroundEar-EEG-based brain-computer interfaces (BCIs) provide improved wearability and comfort compared to traditional scalp-EEG systems. However, their performance is constrained by low signal-to-noise ratios (SNRs) and high rates of BCI illiteracy under conventional luminance-modulated steady-state visual evoked potential (SSVEP) paradigms. MethodsThis study introduces a text-sequence stimulation paradigm to address these limitations by leveraging ventral visual pathway responses that are more accessible to electrodes near the ear. Using offline frequency-sweeping experiments across 4-8 Hz, we identified optimal stimulus parameters (4.6-6.8 Hz with 0.25{pi} phase shifts) and integrated them into a 12-target BCI system. We further conducted online experiments to compare the response characteristics and real-time spelling performance between the proposed text-sequence paradigm and conventional luminance stimulation. ResultsComparative experiments with 14 participants demonstrate that text sequence stimuli achieve an average information transfer rate (ITR) of 44.59 {+/-} 10.50 bits/min, outperforming luminance modulation by 76.18% in ITR. Notably, text sequence stimulation effectively mitigated BCI illiteracy, with all participants achieving near or above 70% accuracy (mean: 86.37 {+/-} 9.61%). This represents a significant improvement over luminance modulation, where 50% of users fell below 70% accuracy. ConclusionsBy reducing the flicker area by 14% and mimicking the natural luminance variations that occur during reading, the proposed method enhanced visual comfort. The online results further validate text-sequence stimulation as a high-performance and user-friendly paradigm for ear-EEG BCIs, supporting their practicality for assistive applications.

This study aims to predict human intentions during intense sports activities, specifically in table tennis. Using a publicly available Real World Table Tennis dataset containing simultaneous EEG and video recordings, we developed a series of participant-specific classifiers for nine players (7 males and 2 females; age range 18-30), based on pose features and EEG signals. The pose-based classifier used a stochastic gradient descent model with logistic loss, whereas the EEG-based classifier employed a modified convolutional neural network architecture (EEGNet). Both classifiers successfully predicted left-right attack intentions from the time windows preceding racket-ball impact, with optimal decoding occurring at -100 ms for pose features and -500 ms for EEG signals. EEG-based decoding achieved higher performance than pose-based decoding, and a multi-modal ensemble further improved prediction, reaching a mean macro F1 score of 0.563 (bootstrapped 95% CI: 0.523-0.603), corresponding to gains of +0.03 over pose-only and +0.02 over EEG-only classifiers. Because each classifier is trained independently, the ensemble can be feasibly extended to incorporate additional modalities in the future. These results suggest potential applications in neural prosthetic systems and neurofeedback tools for sports training.

Electroencephalography (EEG) interpretation in clinical practice relies on the analysis of energy distribution across standard frequency bands. The Weed Plot framework encodes band-wise spectral energy, computed using Fourier-based methods, into a symbolic representation that preserves the interpretability of traditional EEG analysis. In this study, we propose a wavelet-based extension of this framework, where the energy of predefined clinical EEG bands is estimated using the Discrete Wavelet Transform instead of Power Spectral Density. Unlike Fourier-based approaches, wavelets provide a time-frequency representation that captures transient and non-stationary dynamics while remaining consistent with clinically defined bands. From these estimates, symbolic patterns are constructed based on the relative ordering of frequency bands within short temporal windows. Their empirical distribution is used to extract entropy-based features for epilepsy detection using multiple machine learning classifiers. From an Artificial Intelligence perspective, the main contribution is a structured symbolic encoding that enhances feature discriminability. From an engineering perspective, the contribution lies in an automated framework for EEG-based epilepsy detection. Experimental results show that wavelet-based representations improve classification performance compared to raw entropy and Fourier-based features. This improvement arises from the interaction between time-frequency localization and symbolic encoding, producing more discriminative feature distributions. These findings support wavelet-based symbolic representations as a robust and interpretable framework for EEG analysis, bridging clinical interpretation and data-driven methods.

The opportunity to collect movement data from smartphones for prolonged periods has opened new perspectives in the field of clinical movement analysis. However, when monitoring peoples mobility in free-living conditions, smartphone placement can influence the validity of the extracted digital mobility outcome. This study aimed to develop and validate an automatic smartphone placement recognition classifier and to investigate potential critical factors that can influence performance. The classifier was trained on data from 15 healthy participants using inertial signals collected from smartphones placed at six body placements during free-living walking and externally validated on over 3,000 individuals from external datasets, including blind participants and patients with cardiovascular or Parkinsons disease. A decision-tree ensemble model was developed using feature subsets of increasing dimensionality, with the optimal subset comprising 50 features. Classification accuracy increased consistently when front and back pocket placements were aggregated (81.1%) and further improved when coat pocket was also included in the pocket class (88.5%), underscoring the challenge of distinguishing between fine-grained pocket placements. The best-recognized placements across the external datasets were lower back (precision: 100%, recall: 72.5%), hand (precision: 94.2%, recall: 94.5%), and the aggregated pocket class (precision: 86.7%, recall: 90.2%). Recognition accuracy changed across cohorts (0.73 - 0.85), activities (0.63 - 0.94) and speed (0.79 - 0.87), however it stayed consistent across various technological and environmental factors. Overall, this study demonstrates the feasibility of robust placement recognition in walking and underscores the importance of accounting for key influencing factors when designing frameworks intended for deployment in heterogeneous real-world or clinical contexts. HighlightsO_LIMachine learning accurately identifies smartphone placement during real-world gait C_LIO_LISix on-body placements recognized, including pockets, hand, bag, and lower-back C_LIO_LIFree-living data used for training, ensuring robust performance across conditions C_LIO_LIFeature selection and hyperparameter tuning optimize classification accuracy C_LIO_LIExternal validation confirms generalizability across >3,000 healthy and diseased adults C_LI

Distinct smartphone interaction behaviors, like short-form video scrolling and mobile gaming, elicit qualitatively different cognitive and physiological responses. However, such distinctions is often overlooked by approaches that treat smart-phone use as a monolithic behavior. This paper presents Brain-Body Spatio-Temporal Transformer (BB-STT), a unified deep learning framework for classifying interaction-specific physiological signatures from multimodal signals, including EEG, EDA, PPG, and eye-tracking. BB-STT achieves 83.51% accuracy in distinguishing smartphone from non-smartphone activity and 74.13% accuracy in three-class classification of short-form video, gaming, and baseline viewing. The model demonstrates strong generalization with leave-one-subject-out (LOSO) performance that is also comparable to 5-fold cross-validation accuracy. Cross-modal attention emerges as the key component, improving three-class accuracy by 16.74 points through dynamic integration of multimodal signals. Interpretability analysis indicates a hierarchical organization of physiological responses. Eye-tracking features, particularly gaze depth, enable coarse separation between smartphone and non-smartphone activity. In contrast, finer discrimination between passive video viewing and active gaming on smartphones relies on the joint contribution of bilateral pupil dilation and central EEG features. Together, these results demonstrate the potential of multimodal physiological signals for objective, real-time assessment of digital engagement in naturalistic settings.

PurposeHigh-density RF receive arrays are required to realize the inherently available SNR and parallel imaging advantages at ultrahigh field strengths, which are essential for high-resolution functional and anatomical brain MRI. This study aims to systematically assess the impacts of often-overlooked parasitic losses associated with various RF coil components, as these losses can degrade the realized SNR and cause significant deviation from the ultimate intrinsic SNR (uiSNR; the theoretical upper bound of available SNR). In addition, we seek to detail engineering solutions to each of these loss mechanisms in pursuit of achieving a higher fraction of the uiSNR limit. MethodsA 16-channel loop-folded dipole transceiver array was developed for 10.5T human head applications and paired with a fully-updated 64-channel receive-only loop array. The optimization of the receive array considered several factors, including (but not limited to) coil dimensions to accommodate a larger population, the size and number of loops to enhance SNR and parallel imaging performance, and circuit design strategies to minimize parasitic losses. The SNR and parallel imaging performance of the receive array were quantitatively assessed by comparison with the uiSNR, as well as existing high-channel-count receive arrays at 7T and 10.5T. Finally, the complete 16-channel transmit, 80-channel receive coil array was safety validated for human use and employed for high-resolution functional and anatomical MRI at 10.5T. ResultsInitial results show that the 80-channel array, featuring larger loops in an overlapped layout with optimized circuitry, significantly improves the SNR and approaches the uiSNR limit in a large fraction of the head, while maintaining or enhancing the parallel imaging performance compared to previously used non-overlap layout. ConclusionThis study suggests that, although the traditionally used high-channel-count loop receive array technology can approach the uiSNR limit in the >10T regime, meticulous design optimization--including systematic assessment and minimization of parasitic losses--has become increasingly critical for achieving this goal in this new field-strength territory.