Low-latency detection of muscle activation onset is essential for online movement analysis and for many assistive and rehabilitation applications, enabling early inference of user motion intent. Surface electromyography (sEMG) are promising signals for this purpose. However, prior work has highlighted that designing robust online sEMG onset detectors remains challenging due to substantial inter- and intra-subject variability, and many approaches are primarily evaluated offline, limiting their applicability in real-world settings. As a step toward addressing these challenges, we propose an online sEMG onset detector designed for demanding activities of daily living and capable of handling a wide range of signal characteristics. The method is evaluated using a pseudo-online framework providing near real-life testing conditions. The proposed detector relies on the time derivative of fuzzy sample entropy (SampEn) computed on sliding sEMG windows. Validation used two datasets collected for this study: (i) experimental sEMG comprising 1676 samples from 335 sit-to-stand motions recorded from six healthy participants, and (ii) synthetic sEMG signals generated by a physiologically-based model with controlled baseline noise. Evaluation followed a pseudo-online procedure, emulating real-time windowing on streaming data and accounting for filter-induced latency, without prior knowledge of signal amplitude or baseline noise. Performance was compared with two established baseline detectors, respectively based on the Extended Teager-Kaiser Energy Operator (ETKEO) and the fuzzy SampEn, both reimplemented under identical pseudo-online constraints. Metrics included detection delay, missed and false detection rates, and working ranges across signal and noise conditions. On experimental data, muscle activation was detected 36±70 ms before motion onset on average, with pre-motion detection in 77

Monitoring upper-limb activity in daily life is essential for personalizing rehabilitation and assessing functional recovery after stroke. Wearable accelerometers enable continuous and ecological measurement of arm use, but the resulting high-dimensional time-series data remain difficult to interpret in clinical practice. In this work, we propose an interpretable activity profiling framework for post-stroke care based on explainable evidential clustering. Accelerometer signals from both wrists are segmented into five-minute windows and transformed into clinically meaningful features describing functional arm use. Two clustering pipelines are investigated: (i) Evidential C-Means followed by an interpretable decision-tree approximation using Iterative Evidential Mistakeness Minimization (IEMM), and (ii) K-means followed by a Belief Rule-Based Classification System (BRBCS) inspired from Fuzzy Rule-Based Classification Systems (FRBCS). Both approaches produce transparent explanations of cluster assignments while explicitly modeling uncertainty. Clustering results are combined with patient-reported activity logs using evidence theory to derive activity-level profiles that characterize arm use patterns and associated uncertainty. Experiments on real-world post-stroke data show that the proposed framework provides clinically interpretable insights into patients’ behavior across daily activities, supporting individualized rehabilitation planning while preserving robustness to noisy and uncertain data.

Abstract Objective Magnetophosphenes are visual percepts induced by extremely low-frequency magnetic fields (ELF-MF; <300 Hz), yet the EEG expression of suprathreshold magnetophosphene-inducing stimulation remains poorly characterized and is not reliably captured by classical low-frequency markers. We tested whether suprathreshold 20 Hz tAMS is accompanied by broadband high-frequency EEG changes rather than focal oscillatory effects. Approach EEG was recorded in N=13 healthy volunteers during 20 Hz sinusoidal magnetic-field exposure delivered using transcranial alternating magnetic stimulation (tAMS) in a global-head configuration. Three conditions were analyzed: no exposure (0 mT), subthreshold (5 mT), and suprathreshold (50 mT). Gamma-band activity (30–80 Hz) was quantified using complementary spectral approaches, including aperiodic-adjusted gamma power. Main results Perception reports sharply dissociated the three conditions, with frequent perception at 50 mT only. Suprathreshold stimulation was associated with spatially distributed increases in gamma-band activity over frontal and occipito-parietal electrodes. These effects persisted after aperiodic correction using two independent parameterization methods and did not exhibit a consistent narrowband peak, indicating broadband high-frequency changes. No focal power change over primary occipital electrodes remained significant after Bonferroni correction. Significance Suprathreshold magnetophosphene-inducing stimulation is not reliably captured by focal low-frequency EEG markers but is accompanied by distributed broadband high-frequency activity. Because stimulation intensity and perceptual reports were strongly coupled, these effects should be interpreted as EEG correlates of a suprathreshold stimulation-perception state rather than as isolated markers of perception.

Background According to the “loss of complexity” theory, aging and disease are expected to reduce complexity of physiological outputs, thereby limiting the system’s adaptability. However, it remains unclear whether this concept applies to the neuromuscular system in people with chronic obstructive pulmonary disease (pwCOPD). This study aimed to challenge the loss of complexity hypothesis by assessing the regularity, as well as the steadiness and the accuracy, of force production during submaximal isometric contractions in pwCOPD compared to healthy individuals. Methods Seventeen pwCOPD and seventeen age- and sex-matched healthy participants performed submaximal isometric contractions of the knee extensors at six target forces, ranging from 10 to 60% of their maximal voluntary contraction (MVC). Regularity of force signals was assessed using sample entropy (SampEn) and percentage of determinism (DET) from the recurrence quantification analysis. Steadiness and accuracy were quantified using the coefficient of variation (CV) and the root-mean-square error (RMSE), respectively. Results PwCOPD exhibited 26.5% lower MVC than healthy individuals. Despite this muscular weakness, no significant main effect of group or interaction effect (group × contraction intensity) was observed for SampEn, DET, CV and RMSE, suggesting a preserved force control in pwCOPD at all assessed force levels. Conclusion Our results indicate that the loss of complexity theory may not apply in moderate COPD, at least for the neuromuscular system. These findings suggest that neuromuscular alteration associated with COPD may not be sufficient to impair the complexity of force output, questioning the universality of the loss of complexity theory. ### Competing Interest Statement The authors have declared no competing interest. * BMI : Body mass index COPD : Chronic obstructive pulmonary disease CV : Coefficient of variation DET : Percentage of determinism EMD : Empirical mode decomposition FEV1 : Forced expiratory volume in the first second LMM : Linear mixed models MD : Mean difference MMSE : Mini-mental state examination MVC : Maximal voluntary contraction MVF : Maximal voluntary force PwCOPD : People with chronic obstructive pulmonary disease RMSE : Root-mean-square error ROF : Rating-of-fatigue scale RP : Reccurence plot RQA : Recurrence quantification analysis SampEn : Sample entropy SD : Standard deviation Ministère de l’Enseignement Supérieur et de la Recherche, https://ror.org/03sjk9a61 Centre Hospitalier Intercommunal Toulon-La Seyne-sur-Mer, https://ror.org/04wqvjr21

Unsupervised classification is a fundamental machine learning problem. Real-world data often contain imperfections, characterized by uncertainty and imprecision, which are not well handled by traditional methods. Evidential clustering, based on Dempster-Shafer theory, addresses these challenges. This paper explores the underexplored problem of explaining evidential clustering results, which is crucial for high-stakes domains such as healthcare. Our analysis shows that, in the general case, representativity is a necessary and sufficient condition for decision trees to serve as abductive explainers. Building on the concept of representativity, we generalize this idea to accommodate partial labeling through utility functions. These functions enable the representation of "tolerable" mistakes, leading to the definition of evidential mistakeness as explanation cost and the construction of explainers tailored to evidential classifiers. Finally, we propose the Iterative Evidential Mistake Minimization (IEMM) algorithm, which provides interpretable and cautious decision tree explanations for evidential clustering functions. We validate the proposed algorithm on synthetic and real-world data. Taking into account the decision-maker's preferences, we were able to provide an explanation that was satisfactory up to 93

This paper presents a novel framework for realtime human action recognition in industrial contexts, using standard 2D cameras. We introduce a complete pipeline for robust and real-time estimation of human joint kinematics, input to a temporally smoothed Transformer-based network, for action recognition. We rely on a new dataset including 11 subjects performing various actions, to evaluate our approach. Unlike most of the literature that relies on joint center positions (JCP) and is offline, ours uses biomechanical prior, eg. joint angles, for fast and robust real-time recognition. Besides, joint angles make the proposed method agnostic to sensor and subject poses as well as to anthropometric differences, and ensure robustness across environments and subjects. Our proposed learning model outperforms the best baseline model, running also in real-time, along various metrics. It achieves 88 % accuracy and shows great generalization ability, for subjects not facing the cameras. Finally, we demonstrate the robustness and usefulness of our technique, through an online interaction experiment, with a simulated robot controlled in real-time via the recognized actions.

The physical proximity of genomic sites to each other, within the 3D structure of a chromosome, can be experimentally measured in living cells and represented in the form of a matrix, called a contact map. These maps are very similar to recurrence plots of dynamical systems. Consequently, the same methods can be used for reconstructing the underlying spatial structure, either the chromosome 3D spatial structure or the attractor in the phase space. These methods involve two steps: (1) deriving a complete distance matrix of the structure using graph distances on a contact network associated with the contact map; (2) reconstructing the structure from this distance matrix, using well-established methods from distance geometry or multidimensional scaling. We here review the different options for implementing the first step, according to the binary, graded or weighted nature of the contact map. We illustrate on three benchmarks (Lorenz model, white noise and EEG data) a novel spectral criterion reflecting the spatial dimension relevant in the second step.

Dystrophin deficiency alters the sarcolemma structure, leading to muscle dystrophy, muscle disuse, and ultimately death. Beyond limb muscle deficits, patients with Duchenne muscular dystrophy have numerous transit disorders. Many studies have highlighted the strong relationship between gut microbiota and skeletal muscle. The aims of this study were: i) to characterize the gut microbiota composition over time up to 1 year in dystrophin-deficient mdx mice, and ii) to analyze the intestine structure and function and expression of genes linked to bacterial-derived metabolites in ileum, blood, and skeletal muscles to study interorgan interactions. Mdx mice displayed a significant reduction in the overall number of different operational taxonomic units and their abundance (α-diversity). Mdx genotype predicted 20% of β-diversity divergence, with a large taxonomic modification of Actinobacteria, Proteobacteria, Tenericutes, and Deferribacteres phyla and the included genera. Interestingly, mdx intestinal motility and gene expressions of tight junction and Ffar2 receptor were down-regulated in the ileum. Concomitantly, circulating inflammatory markers related to gut microbiota (tumor necrosis factor, IL-6, monocyte chemoattractant protein-1) and muscle inflammation Tlr4/Myd88 pathway (Toll-like receptor 4, which recognizes pathogen-associated molecular patterns) were up-regulated. Finally, in mdx mice, adiponectin was reduced in blood and its receptor modulated in muscles. This study highlights a specific gut microbiota composition and highlights interorgan interactions in mdx physiopathology with gut microbiota as the potential central metabolic organ.

Correlation coefficients play a pivotal role in quantifying linear relationships between random variables. Yet, their application to time series data is very challenging due to temporal dependencies. This paper introduces a novel approach to estimate the statistical significance of correlation coefficients in time series data, addressing the limitations of traditional methods based on the concept of effective degrees of freedom (or effective sample size, ESS). These effective degrees of freedom represent the independent sample size that would yield comparable test statistics under the assumption of no temporal correlation. We propose to assume a parametric Gaussian form for the autocorrelation function. We show that this assumption, motivated by a Laplace approximation, enables a simple estimator of the ESS that depends only on the temporal derivatives of the time series. Through numerical experiments, we show that the proposed approach yields accurate statistics while significantly reducing computational complexity, from O(nlogn) to O(n). In addition, we evaluate the adequacy of our approach on real physiological signals, for assessing the connectivity measures in electrophysiology and detecting correlated arm movements in motion capture data. Our methodology provides a simple tool for researchers working with time series data, enabling robust hypothesis testing in the presence of temporal dependencies.

Recent studies suggest that, compared to healthy individuals, people with chronic obstructive pulmonary disease (pwCOPD) present a reduced capacity to perform cognitive-motor dual-task (CMDT). However, these studies were focused on short-duration CMDT offering limited insight to prolonged CMDT inducing fatigue, which can be encountered in daily life. The present study aimed to explore the effect of adding a cognitive task during repeated muscle contractions on muscle endurance, neuromuscular fatigability and cognitive control in pwCOPD compared to healthy participants. Thirteen pwCOPD and thirteen age- and sex-matched healthy participants performed submaximal isometric contractions of the knee extensors until exhaustion in two experimental sessions: (1) without cognitive task and (2) with a concurrent working memory task (i.e., 1-back task). Neuromuscular fatigability (as well as central and peripheral components measured by peripheral magnetic stimulation), cognitive performance and perceived muscle fatigue were assessed throughout the fatiguing tasks. Independently to the experimental condition, pwCOPD exhibited lower muscle endurance compared to healthy participants (p=0.039), mainly explained by earlier peripheral fatigue and faster attainment of higher perceived muscle fatigue (p<0.05). However, neither effect of cognitive task (p=0.223) nor interaction effect (group × condition ; p=0.136) was revealed for muscle endurance. Interestingly, cognitive control was significantly reduced only in pwCOPD at the end of CMDT (p<0.015), suggesting greater difficulty for patients with dual-tasking under fatigue. These findings provide novel insights into how and why fatigue develops in COPD in dual-task context, offering a rationale for including such tasks in rehabilitation programs.

et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Background: Electric vestibular stimulations (EVS) up to 300 Hz trigger vestibular myogenic responses. Interestingly, 300 Hz is the upper limit of the so called extremely low-frequency magnetic fields (ELF-MF) range found within the 2010 guidelines written by the International Commission for Non-Ionizing Radiation Protection. Such guidelines are used to protect the workers and the public from neurostimulation triggered by induced electric fields. Since EVS is known to bias reaching and pointing tasks, vestibular-specific electric fields at power-line frequency are likely to impact the safety and performance of workers in high ELF-MF environments. Objectives: This research aimed to investigate the impact of vestibular-specific electric-fields on manual pointing accuracy. Methods: Pointing accuracy of twenty healthy participants was analyzed with both direct current (2 mA) and sinusoidal (peak ± 2 mA at 50 Hz) EVS. Spatial orientation and quantity of movement variables were used to investigate pointing modulations. Results: Despite a pre-trial conclusive positive control effect, no significant effects of both direct current and 50 Hz stimulation exposures were found. Conclusions: Although high vestibular-specific electric fields were used; no pointing accuracy modulation was found. These results suggest that ELF exposure even at high levels are not able to modulate hand pointing performance in humans. Even though this could be explained by context-specific habituation mechanisms rapidly decreasing EVS impact over time, these results represent useful knowledge for the safety and the performance of workers evolving in high ELF-MF environments.

FixedMotion Capture systems (Mocap) are wellestablished nowadays and are available in indoor and outdoor environments. Here, we introduce RobCap, a Moving Motion Capture System designed for acquiring accurate three-dimensional kinematics of human motion with large displacements. RobCap consists of an optoelectronicMocap (Codamotion) mounted on a robotic arm, to track the human motion in the local moving frame. This paper presents this device and expresses the local kinematics of the markers in a global coordinate system. We have acquired a set of 560 measurements, by controlling the Robotic arm to follow the subject's movement and to track the markers attached to his wrist. The performance of the RobCap is evaluated through comparison with state-of-the-art technologies. The 3D error is found to be 9.7 +/- 4.1 mm and 3.9 +/- 0.8 mm, respectively without and with calibration. RobCap increases the nominal capture volume of the fixedMocap from 75 m(3) to 523m(3) and the operational volume from105m(3) to 1436m(3). Our findings show that the RobCap is accurate and precise enough to be used as a mobile Mocap system for large displacements.

Intuitive user interfaces are indispensable to interact with the human centric smart environments. In this paper, we propose a unified framework that recognizes both static and dynamic gestures, using simple RGB vision (without depth sensing). This feature makes it suitable for inexpensive human-robot interaction in social or industrial settings. We employ a pose-driven spatial attention strategy, which guides our proposed Static and Dynamic gestures Network-StaDNet. From the image of the human upper body, we estimate his/her depth, along with the region-of-interest around his/her hands. The Convolutional Neural Network (CNN) in StaDNet is fine-tuned on a background-substituted hand gestures dataset. It is utilized to detect 10 static gestures for each hand as well as to obtain the hand image-embeddings. These are subsequently fused with the augmented pose vector and then passed to the stacked Long Short-Term Memory blocks. Thus, human-centred frame-wise information from the augmented pose vector and from the left/right hands image-embeddings are aggregated in time to predict the dynamic gestures of the performing person. In a number of experiments, we show that the proposed approach surpasses the state-of-the-art results on the large-scale Chalearn 2016 dataset. Moreover, we transfer the knowledge learned through the proposed methodology to the Praxis gestures dataset, and the obtained results also outscore the state-of-the-art on this dataset.

Recurrence quantification analysis (RQA) is a nonlinear method providing information on the temporal structure of time series. RQA has been extensively used to explore various noisy and nonstationary physiological signals. However, the application of RQA to force signals acquired during voluntary fatiguing contractions performed until exhaustion remain to be investigated. We aimed to explore the sensitivity of the percentage of determinism (DET), an RQA predictability measure, to detect changes of force signal complexity induced by fatigue and recovery. Changes in force signal complexity were compared between women and men to explore the ability of DET measures to detect different fatigue profiles. Nineteen women and nineteen men performed intermittent isometric contractions of knee extensors at 50% of maximal voluntary contraction (MVC) until exhaustion. Participants performed MVC before, during and after the fatiguing task to assess neuromuscular fatigue. Recovery measurements were performed three minutes after exhaustion. Particular attention has been given to the selection of the input parameters of RQA and to the influence of nonstationarity. A detailed methodology is provided to apply RQA to force signals. At the whole group level, complexity decreased with fatigue then increased after recovery. Greater fatigability of men was associated with a faster loss of complexity (i.e. faster increase of DET) of force signals. After recovery, complexity returned to baseline value only for women. These findings confirm that RQA is suited to explore force signal temporal structure and is able to reveal changes of complexity induced by fatigue and recovery by taking into account sex differences.

Direct electrical stimulation (DES) is used to perform functional brain mapping during awake surgery and in epileptic patients. DES may be coupled with the measurement of Evoked Potentials (EP) to study the conductive and integrative properties of activated neural ensembles and probe the spatiotemporal dynamics of short- and long-range networks. However, its electrophysiological effects remain by far unknown. We recorded ECoG signals on two patients undergoing awake brain surgery and measured EP on functional sites after cortical stimulations and were the firsts to record three different types of EP on the same patients. Using low-intensity (1–3 mA) to evoke electrogenesis we observed that: (i) “true” remote EPs are attenuated in amplitude and delayed in time due to the divergence of white matter pathways; (ii) “false” remote EPs are attenuated but not delayed: as they originate from the same electrical source; (iii) Singular but reproducible positive components in the EP can be generated when the DES is applied in the temporal lobe or the premotor cortex; and (iv) rare EP can be triggered when the DES is applied subcortically: these can be either negative, or surprisingly, positive. We proposed different activation and electrophysiological propagation mechanisms following DES, based on the nature of activated neural elements and discussed important methodological pitfalls when measuring EP in the brain. Altogether, these results pave the way to map the connectivity in real-time between the DES and the recording sites; to characterize the local electrophysiological states and to link electrophysiology and function. In the future, and in practice, this technique could be used to perform electrophysiological mapping in order to link (non)-functional to electrophysiological responses with DES and could be used to guide the surgical act itself.

•Analytic expressions of five RQA measures for autoregressive processes are derived.•Parametric RQA (pRQA) applies to time series modeled by autoregressive processes.•pRQA is computationally fast and accurate.•pRQA can detect spatial patterns in multichannel data, e.g. EEG data.

When nonlinear measures are estimated from sampled temporal signals with finite-length, a radius parameter must be carefully selected to avoid a poor estimation. These measures are generally derived from the correlation integral, which quantifies the probability of finding neighbors, i.e., pair of points spaced by less than the radius parameter. While each nonlinear measure comes with several specific empirical rules to select a radius value, we provide a systematic selection method. We show that the optimal radius for nonlinear measures can be approximated by the optimal bandwidth of a Kernel Density Estimator (KDE) related to the correlation sum. The KDE framework provides non-parametric tools to approximate a density function from finite samples (e.g., histograms) and optimal methods to select a smoothing parameter, the bandwidth (e.g., bin width in histograms). We use results from KDE to derive a closed-form expression for the optimal radius. The latter is used to compute the correlation dimension and to construct recurrence plots yielding an estimate of Kolmogorov-Sinai entropy. We assess our method through numerical experiments on signals generated by nonlinear systems and experimental electroencephalographic time series.

Physiological signals present fluctuations that can be assessed from their temporal structure, also termed complexity. The complexity of a physiological signal is usually quantified using entropy estimators, such as Sample Entropy. Recent studies have shown a loss of force signal complexity with the development of neuromuscular fatigue. However, these studies did not consider the stationarity of the force signals which is an important prerequisite of Sample Entropy measurements. Here, we investigated the effect of the potential nonstationarity of force signals on the kinetics of neuromuscular fatigue-induced change in force signal's complexity. Eleven men performed submaximal intermittent isometric contractions of knee extensors until exhaustion. Neuromuscular fatigue was assessed from changes in voluntary and electrically evoked contractions. Sample Entropy values were computed from submaximal force signals throughout the fatiguing task. The Dickey-Fuller test was used to statistically investigate the stationarity of force signals and the Empirical Mode Decomposition was applied to detrend these signals. Maximal voluntary force, central voluntary activation and muscle twitch decreased throughout the task (all ), indicating the development of global, central and peripheral fatigue, respectively. We found an increase in Sample Entropy with fatigue ( p = 0.024 ) when not considering the nonstationarity of force signals (i.e., 43% of nonstationary signals). After applying the Empirical Mode Decomposition, we found a decrease in Sample Entropy with fatigue ( p = 0.002 ). These findings confirm the presence of nonstationarity in force signals during submaximal isometric contractions which influences the kinetics of Sample Entropy with neuromuscular fatigue.