Pub Date : 2026-01-26DOI: 10.1109/TNSRE.2026.3657810
Xingjian Li;Xiang Chen;Penghui Lin;De Wu;Xu Zhang
Functional electrical stimulation (FES) is commonly used in clinical practice to induce muscle contractions for rehabilitation therapy. However, the collected electromyography (EMG) under FES is a mixed signal containing voluntary electromyography (VEMG), which is generated by neural system recruited motor units, and FES response, which is composed of initial spikes generated by current passage and M-waves generated by stimulation recruited motor units. To realize close-loop control of FES system or to reveal the mechanism of FES through EMG signal, it is necessary to separate the three components from mixed signal. Based on the assumption that signal sources of initial spikes and action potentials related signals (VEMG and M-waves) are independent of each other, this study presents a novel mixed FES signal decomposition algorithm termed FastICA-Comb. Specifically, in order to achieve high-quality separation of VEMG, M-waves and initial spikes, the algorithm is designed as a unique scheme combining one-stage comb filtering and two-stage fast independent component analysis (FastICA) decomposition and classification. To evaluate the proposed algorithm’s effectiveness, FES data collection experiment was conducted on 6 healthy subjects. The experimental results confirm that the FastICA-Comb algorithm has better VEMG and M-wave extraction capabilities than the classical methods including comb filter, GS-PEF, EMD-notch and blanking window in both simulated and real mixed FES signal. Therefore, the research findings provide an effective signal analysis tool for exploring the therapeutic mechanism of FES.
{"title":"FastICA-Comb: A Novel Algorithm for Extracting Voluntary Electromyography and M-Wave in Functional Electrical Stimulation Scenarios","authors":"Xingjian Li;Xiang Chen;Penghui Lin;De Wu;Xu Zhang","doi":"10.1109/TNSRE.2026.3657810","DOIUrl":"10.1109/TNSRE.2026.3657810","url":null,"abstract":"Functional electrical stimulation (FES) is commonly used in clinical practice to induce muscle contractions for rehabilitation therapy. However, the collected electromyography (EMG) under FES is a mixed signal containing voluntary electromyography (VEMG), which is generated by neural system recruited motor units, and FES response, which is composed of initial spikes generated by current passage and M-waves generated by stimulation recruited motor units. To realize close-loop control of FES system or to reveal the mechanism of FES through EMG signal, it is necessary to separate the three components from mixed signal. Based on the assumption that signal sources of initial spikes and action potentials related signals (VEMG and M-waves) are independent of each other, this study presents a novel mixed FES signal decomposition algorithm termed FastICA-Comb. Specifically, in order to achieve high-quality separation of VEMG, M-waves and initial spikes, the algorithm is designed as a unique scheme combining one-stage comb filtering and two-stage fast independent component analysis (FastICA) decomposition and classification. To evaluate the proposed algorithm’s effectiveness, FES data collection experiment was conducted on 6 healthy subjects. The experimental results confirm that the FastICA-Comb algorithm has better VEMG and M-wave extraction capabilities than the classical methods including comb filter, GS-PEF, EMD-notch and blanking window in both simulated and real mixed FES signal. Therefore, the research findings provide an effective signal analysis tool for exploring the therapeutic mechanism of FES.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"918-929"},"PeriodicalIF":5.2,"publicationDate":"2026-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11363479","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146051945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-23DOI: 10.1109/TNSRE.2026.3657400
Dario Di Domenico;Fabio Egle;Andrea Marinelli;Michele Canepa;Emanuele Gruppioni;Nicoló Boccardo;Matteo Laffranchi;Claudio Castellini
Upper-limb prosthesis control remains challenging in achieving natural and intuitive movements, especially for devices with multiple actuated degrees of freedom (DoFs), often demanding high cognitive effort. Machine learning aids in mapping phantom limb muscle patterns to prosthetic movements, but is limited by the instability of electromyographic signals over time. This study investigates two simultaneous and proportional myocontrol strategies, based on position and velocity, using incremental learning for a 3-DoFs prosthesis, allowing co-adaptation between the system and the user. Six able-bodied and five limb-difference participants performed Target Achievement Control tests over four sessions per control strategy, assessing performance, usability, workload, simultaneity, and proportionality. Results indicate that velocity control consistently outperforms position control in both populations, yielding lower errors, higher success rates and path efficiency, and lower workload. Notably, both control strategies showed significant improvement over time in the able-bodied group, while only position control improved significantly in the limb-difference group. Interestingly, no significant difference in usability was observed between the two strategies in either group. Position control promoted greater simultaneous actuation of multi-DoFs. However, the overall findings support the use of velocity-based control as a means to improve prosthetic performance and user satisfaction.
{"title":"Co-Adaptive Velocity and Position Control of 3-DoFs Prosthesis via Incremental Learning","authors":"Dario Di Domenico;Fabio Egle;Andrea Marinelli;Michele Canepa;Emanuele Gruppioni;Nicoló Boccardo;Matteo Laffranchi;Claudio Castellini","doi":"10.1109/TNSRE.2026.3657400","DOIUrl":"10.1109/TNSRE.2026.3657400","url":null,"abstract":"Upper-limb prosthesis control remains challenging in achieving natural and intuitive movements, especially for devices with multiple actuated degrees of freedom (DoFs), often demanding high cognitive effort. Machine learning aids in mapping phantom limb muscle patterns to prosthetic movements, but is limited by the instability of electromyographic signals over time. This study investigates two simultaneous and proportional myocontrol strategies, based on position and velocity, using incremental learning for a 3-DoFs prosthesis, allowing co-adaptation between the system and the user. Six able-bodied and five limb-difference participants performed Target Achievement Control tests over four sessions per control strategy, assessing performance, usability, workload, simultaneity, and proportionality. Results indicate that velocity control consistently outperforms position control in both populations, yielding lower errors, higher success rates and path efficiency, and lower workload. Notably, both control strategies showed significant improvement over time in the able-bodied group, while only position control improved significantly in the limb-difference group. Interestingly, no significant difference in usability was observed between the two strategies in either group. Position control promoted greater simultaneous actuation of multi-DoFs. However, the overall findings support the use of velocity-based control as a means to improve prosthetic performance and user satisfaction.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"856-867"},"PeriodicalIF":5.2,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11363318","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146040767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-23DOI: 10.1109/TNSRE.2026.3657614
Xuna Wang;Hongwei Gao;Yanxin Cui;Jiahui Yu;Gongfa Li;Zhaojie Ju
As skeletal data can be collected non-invasively while preserving patient privacy, it is widely used in public medical datasets to document patient behavior. Autism Spectrum Disorder (ASD) is characterized by significant behavioral heterogeneity, reflected in the topological structure and dynamic evolution of skeletal movements. This complexity poses substantial challenges for skeleton-based behavioral analysis. Existing methods struggle to effectively utilize behavioral evolution for subject-specific reasoning, leading to suboptimal representations that lack diagnostic relevance for autism. To address this limitation, we propose a Behavioral Evolution-based Edge Reconstruction (BER) Strategy for learning autism-related behavioral representations. By reconstructing a high-granularity adjacency matrix that spans both spatial and temporal dimensions, utilizing dynamic evolution and spatial location information, BERGCN enhances behavioral reasoning. Specifically, we first compute channel-level spatial and temporal edge reconstruction parameters by performing feature compression and targeted convolution operations on the differences between neighboring frames. Based on these, the spatial edge reconstruction module is designed by combining a generic attention map with two personalized attention maps, while the temporal edge reconstruction module is implemented using flexible frame replacement and weighted aggregation. Finally, we investigate both single-modal and multimodal network architectures under various fusion strategies. We evaluate BERGCN on three autism clinical tasks and a benchmark action recognition dataset. Experimental results demonstrate competitive performance, showing improved sensitivity to subject-specific behavioral patterns while maintaining computational efficiency.
{"title":"Modeling the Heterogeneous Movements of ASD via Fine-Grained Skeleton Representation Learning","authors":"Xuna Wang;Hongwei Gao;Yanxin Cui;Jiahui Yu;Gongfa Li;Zhaojie Ju","doi":"10.1109/TNSRE.2026.3657614","DOIUrl":"10.1109/TNSRE.2026.3657614","url":null,"abstract":"As skeletal data can be collected non-invasively while preserving patient privacy, it is widely used in public medical datasets to document patient behavior. Autism Spectrum Disorder (ASD) is characterized by significant behavioral heterogeneity, reflected in the topological structure and dynamic evolution of skeletal movements. This complexity poses substantial challenges for skeleton-based behavioral analysis. Existing methods struggle to effectively utilize behavioral evolution for subject-specific reasoning, leading to suboptimal representations that lack diagnostic relevance for autism. To address this limitation, we propose a Behavioral Evolution-based Edge Reconstruction (BER) Strategy for learning autism-related behavioral representations. By reconstructing a high-granularity adjacency matrix that spans both spatial and temporal dimensions, utilizing dynamic evolution and spatial location information, BERGCN enhances behavioral reasoning. Specifically, we first compute channel-level spatial and temporal edge reconstruction parameters by performing feature compression and targeted convolution operations on the differences between neighboring frames. Based on these, the spatial edge reconstruction module is designed by combining a generic attention map with two personalized attention maps, while the temporal edge reconstruction module is implemented using flexible frame replacement and weighted aggregation. Finally, we investigate both single-modal and multimodal network architectures under various fusion strategies. We evaluate BERGCN on three autism clinical tasks and a benchmark action recognition dataset. Experimental results demonstrate competitive performance, showing improved sensitivity to subject-specific behavioral patterns while maintaining computational efficiency.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"1021-1030"},"PeriodicalIF":5.2,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11363308","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146040754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-23DOI: 10.1109/TNSRE.2026.3657728
Andrea Scarpellini;Lorenzo Di Silverio;Anna Carroll;Leora R. Cherney;Edna M. Babbitt;James L. Patton;Hananeh Esmailbeigi
The tongue is a uniquely agile muscular structure essential for vital tasks of speech, breathing, chewing, and swallowing, functions commonly disrupted following neurological injury. Yet, current rehabilitation approaches lack objective measures and techniques to characterize impairment and restore the tongue’s ability. Here, we introduce a clinic-friendly method that isolates and quantifies tongue agility, defined as the ability to execute rapid and precise movements, using a wireless intraoral sensing device that provides real-time visual feedback of movement. Six participants diagnosed with dysarthria completed seven one-hour intervention sessions. Tongue movement probability distributions were generated to identify individualized deviations from neurotypical patterns. An individualized visual feedback intervention was designed to redirect movement away from over-expressed regions toward under-expressed deficient areas. Across the intervention, sensing area coverage increased significantly by 10.29 %, while over-expressed areas decreased significantly by 3.99 %, and movement velocity improved significantly by 3.85 %. This pilot study provides promising preliminary evidence that precision visual feedback rehabilitation can reshape tongue movement patterns and enhance tongue agility in individuals with oral motor disorders.
{"title":"Tongue-Yoga: Precision Visual Feedback Rehabilitation Improves Tongue Agility","authors":"Andrea Scarpellini;Lorenzo Di Silverio;Anna Carroll;Leora R. Cherney;Edna M. Babbitt;James L. Patton;Hananeh Esmailbeigi","doi":"10.1109/TNSRE.2026.3657728","DOIUrl":"10.1109/TNSRE.2026.3657728","url":null,"abstract":"The tongue is a uniquely agile muscular structure essential for vital tasks of speech, breathing, chewing, and swallowing, functions commonly disrupted following neurological injury. Yet, current rehabilitation approaches lack objective measures and techniques to characterize impairment and restore the tongue’s ability. Here, we introduce a clinic-friendly method that isolates and quantifies tongue agility, defined as the ability to execute rapid and precise movements, using a wireless intraoral sensing device that provides real-time visual feedback of movement. Six participants diagnosed with dysarthria completed seven one-hour intervention sessions. Tongue movement probability distributions were generated to identify individualized deviations from neurotypical patterns. An individualized visual feedback intervention was designed to redirect movement away from over-expressed regions toward under-expressed deficient areas. Across the intervention, sensing area coverage increased significantly by 10.29 %, while over-expressed areas decreased significantly by 3.99 %, and movement velocity improved significantly by 3.85 %. This pilot study provides promising preliminary evidence that precision visual feedback rehabilitation can reshape tongue movement patterns and enhance tongue agility in individuals with oral motor disorders.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"1352-1362"},"PeriodicalIF":5.2,"publicationDate":"2026-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11363165","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146040759","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-19DOI: 10.1109/TNSRE.2026.3656061
Jacob Salminen;Chang Liu;Erika M. Pliner;Arkaprava Roy;Natalie Richer;Jungyun Hwang;Chris J. Hass;David J. Clark;Yenisel Cruz-Almeida;Todd M. Manini;Rachael D. Seidler;Daniel P. Ferris
Aging alters both biomechanical and neural factors related to walking, leading to reductions in preferred gait speed with age. Biomechanical variability in human walking has been an area of great interest for aging research. Neural variability has not been well studied in this context. Electroencephalography (EEG) can measure brain activity during walking, allowing us to quantify interstride variability of electrocortical activity. We recruited younger and older adults to walk (0.25-1.0 m/s) while we measured EEG interstride variability in theta, alpha, and beta power. We hypothesized that theta, alpha, and beta variability would decrease at faster walking speeds like most gait kinematic variables. We also hypothesized that older adults would have more interstride variability compared to younger due to reduced gait automaticity. We observed sensorimotor and posterior parietal cortices for their roles in motor action and sensory processing. Interstride variability in theta power lessened with faster walking speeds in posterior parietal cortex, and Interstride variability in alpha and beta power lessened in both sensorimotor and posterior parietal cortex. Further, we found that older adults had less interstride variability than younger adults, primarily in alpha and beta. We also observed interstride phasic alignment of electrocortical activity across the gait cycle. We found broadband increases in interstride phase alignment across the gait cycle, and that older higher functioning adults had greater phase alignment in gamma (30-50 Hz) than younger adults in parietal cortex. These findings suggest that the automaticity of gait is greater at faster walking speeds, and that older adults’ reduced automaticity of gait may be unrelated to electrical brain activity.
{"title":"Interstride Variation in EEG Power Spectra of Younger and Older Adults Walking at a Range of Gait Speeds","authors":"Jacob Salminen;Chang Liu;Erika M. Pliner;Arkaprava Roy;Natalie Richer;Jungyun Hwang;Chris J. Hass;David J. Clark;Yenisel Cruz-Almeida;Todd M. Manini;Rachael D. Seidler;Daniel P. Ferris","doi":"10.1109/TNSRE.2026.3656061","DOIUrl":"10.1109/TNSRE.2026.3656061","url":null,"abstract":"Aging alters both biomechanical and neural factors related to walking, leading to reductions in preferred gait speed with age. Biomechanical variability in human walking has been an area of great interest for aging research. Neural variability has not been well studied in this context. Electroencephalography (EEG) can measure brain activity during walking, allowing us to quantify interstride variability of electrocortical activity. We recruited younger and older adults to walk (0.25-1.0 m/s) while we measured EEG interstride variability in theta, alpha, and beta power. We hypothesized that theta, alpha, and beta variability would decrease at faster walking speeds like most gait kinematic variables. We also hypothesized that older adults would have more interstride variability compared to younger due to reduced gait automaticity. We observed sensorimotor and posterior parietal cortices for their roles in motor action and sensory processing. Interstride variability in theta power lessened with faster walking speeds in posterior parietal cortex, and Interstride variability in alpha and beta power lessened in both sensorimotor and posterior parietal cortex. Further, we found that older adults had less interstride variability than younger adults, primarily in alpha and beta. We also observed interstride phasic alignment of electrocortical activity across the gait cycle. We found broadband increases in interstride phase alignment across the gait cycle, and that older higher functioning adults had greater phase alignment in gamma (30-50 Hz) than younger adults in parietal cortex. These findings suggest that the automaticity of gait is greater at faster walking speeds, and that older adults’ reduced automaticity of gait may be unrelated to electrical brain activity.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"952-965"},"PeriodicalIF":5.2,"publicationDate":"2026-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11358987","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146003315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-19DOI: 10.1109/TNSRE.2026.3654935
Hyunmi Lim;Hyoseon Choi;Bilal Ahmed;Yoonghil Park;Jeonghun Ku
Stroke frequently results in long-term motor deficits that impair quality of life. Action observation therapy (AOT) has shown promise for motor recovery through engagement of the mirror neuron system (MNS), yet its passive nature and lack of attentional tracking limit its neuroplasticity efficacy. To address these limitations, we developed a closed-loop Brain-Computer Interface-integrated AOT (BCI-AOT) system employing real-time Steady-State Visual Evoked Potential (SSVEP)-based attention monitoring to dynamically control therapy delivery. In a within-subject crossover study, 22 individuals with hemiplegic stroke completed both BCI-AOT and conventional AOT conditions, each consisting of five daily sessions and separated by a one-week washout. In BCI-AOT, video playback depended on sustained attentional engagement detected via SSVEPs. Behavioral outcomes (Box and Block Test [BBT], Action Research Arm Test [ARAT]) and physiological measures (Motor Evoked Potential [MEP] amplitude and latency, EEG power) were assessed pre- and post-intervention. Significant Condition $times $ Day interactions were found for both BBT and ARAT, indicating greater functional gains over time in the BCI-AOT condition. Both conditions showed increased corticospinal excitability over time, while BCI-AOT additionally exhibited distinct mu and theta modulation over time. Participants also reported greater motivation and attention after BCI-AOT compared to conventional AOT. These results suggest that BCI-AOT elicits stronger neuroplasticity responses and user engagement than standard AOT. This study supports the feasibility and clinical potential of closed-loop, attention-adaptive neurorehabilitation for stroke recovery.
中风经常导致长期运动障碍，从而影响生活质量。动作观察疗法（AOT）通过参与镜像神经元系统（MNS）显示出运动恢复的希望，但其被动性质和缺乏注意跟踪限制了其神经可塑性的效果。为了解决这些限制，我们开发了一种闭环脑机接口集成AOT （BCI-AOT）系统，采用基于实时稳态视觉诱发电位（SSVEP）的注意力监测来动态控制治疗递送。在一项受试者内交叉研究中，22名偏瘫中风患者完成了BCI-AOT和常规AOT条件，每个条件由每天五个疗程组成，间隔一周的洗脱期。在BCI-AOT中，视频播放依赖于通过ssvep检测到的持续注意力投入。评估干预前后的行为结果（Box and Block Test [BBT]、动作研究臂测试[ARAT]）和生理指标（运动诱发电位[MEP]振幅、潜伏期、脑电图功率）。在BBT和ARAT中发现了显著的条件×日相互作用，表明BCI-AOT条件下随着时间的推移功能增加更大。随着时间的推移，这两种情况都显示出皮质脊髓兴奋性的增加，而BCI-AOT也表现出明显的mu和theta调制。与常规AOT相比，BCI-AOT后参与者也报告了更大的动机和注意力。这些结果表明BCI-AOT比标准AOT引起更强的神经可塑性反应和用户参与。本研究支持闭环、注意适应性神经康复治疗脑卒中康复的可行性和临床潜力。
{"title":"Attention-Adaptive BCI-AOT System Enhances Motor Recovery and Neural Engagement After Stroke","authors":"Hyunmi Lim;Hyoseon Choi;Bilal Ahmed;Yoonghil Park;Jeonghun Ku","doi":"10.1109/TNSRE.2026.3654935","DOIUrl":"10.1109/TNSRE.2026.3654935","url":null,"abstract":"Stroke frequently results in long-term motor deficits that impair quality of life. Action observation therapy (AOT) has shown promise for motor recovery through engagement of the mirror neuron system (MNS), yet its passive nature and lack of attentional tracking limit its neuroplasticity efficacy. To address these limitations, we developed a closed-loop Brain-Computer Interface-integrated AOT (BCI-AOT) system employing real-time Steady-State Visual Evoked Potential (SSVEP)-based attention monitoring to dynamically control therapy delivery. In a within-subject crossover study, 22 individuals with hemiplegic stroke completed both BCI-AOT and conventional AOT conditions, each consisting of five daily sessions and separated by a one-week washout. In BCI-AOT, video playback depended on sustained attentional engagement detected via SSVEPs. Behavioral outcomes (Box and Block Test [BBT], Action Research Arm Test [ARAT]) and physiological measures (Motor Evoked Potential [MEP] amplitude and latency, EEG power) were assessed pre- and post-intervention. Significant Condition <inline-formula> <tex-math>$times $ </tex-math></inline-formula> Day interactions were found for both BBT and ARAT, indicating greater functional gains over time in the BCI-AOT condition. Both conditions showed increased corticospinal excitability over time, while BCI-AOT additionally exhibited distinct mu and theta modulation over time. Participants also reported greater motivation and attention after BCI-AOT compared to conventional AOT. These results suggest that BCI-AOT elicits stronger neuroplasticity responses and user engagement than standard AOT. This study supports the feasibility and clinical potential of closed-loop, attention-adaptive neurorehabilitation for stroke recovery.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"686-698"},"PeriodicalIF":5.2,"publicationDate":"2026-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11358787","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146003352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-19DOI: 10.1109/TNSRE.2026.3655791
Michelle van Mierlo;Katrijn Smulders;Noël Keijsers
Daily life gait performance measures can provide ecologically valid gait characteristics, which are interesting for monitoring individuals with gait impairments. The first step in obtaining these gait characteristics is selecting walking periods from multiple day recordings. We developed and validated an algorithm for walking detection using inertial measurement units (IMU) on Both feet and compared the performance with two others in healthy individuals and those with neurologically impaired gait: 1) using Sacrum accelerometer data (Iluz et al. 2014); 2) using Single foot gyroscopic data (Ullrich et al. 2020). We also assessed which activities reduce the algorithms’ specificity for walking detection. The Both feet algorithm consisted of three stages, 1) selecting time periods potentially containing walking; 2) excluding periods not containing walking; 3) checking the selected periods for minimal walking bout requirements. For validation, 32 participants (12 healthy and 20 with neurologically impaired gait) performed 20-30 minutes of daily life activities, while wearing IMUs on both feet and the sacrum. Using labelled video recordings as reference, we calculated each algorithm’s specificity, sensitivity and accuracy for walking detection. Both feet outperformed the other algorithms on specificity (96.6% versus 92.1% and 72.1% for the Single foot en Sacrum respectively). Stair climbing was misclassified as walking most often by all algorithms. Sacrum outperformed the others on sensitivity (99.5%), but had low specificity and accuracy. The high specificity of the Both feet algorithm makes it suitable when spatiotemporal gait characteristics are of interest, and is applicable in populations with mild neurological conditions affecting gait.
日常生活步态性能测量可以提供生态有效的步态特征，这对监测步态障碍的个体很有意义。获得这些步态特征的第一步是从多天记录中选择步行周期。我们开发并验证了一种使用双脚惯性测量单元（IMU）进行行走检测的算法，并将其与健康个体和神经系统步态受损者的其他两种算法进行了比较：1)使用骶骨加速度计数据（Iluz et al. 2014）；2)使用单脚陀螺仪数据（Ullrich et al. 2020）。我们还评估了哪些活动降低了算法对步行检测的特异性。双脚算法包括三个阶段：1)选择可能包含步行的时间段；2)不包括散步时段；3)检查所选时段的最小步行回合要求。为了验证，32名参与者（12名健康参与者和20名神经性步态受损参与者）在双脚和骶骨上佩戴imu，进行20- 30分钟的日常生活活动。以标记视频为参考，我们计算了每种算法在行走检测中的特异性、灵敏度和准确性。两只脚的特异性优于其他算法（分别为96.6%和92.1%，单脚和骶骨分别为72.1%）。爬楼梯最常被所有算法错误地归类为步行。骶骨的敏感性为99.5%，但特异性和准确性较低。双脚算法的高特异性使其适用于对时空步态特征感兴趣的情况，并且适用于轻度神经系统疾病影响步态的人群。
{"title":"Feet-Based IMU Algorithm Yields High Specificity for Detection of Walking in Daily Life","authors":"Michelle van Mierlo;Katrijn Smulders;Noël Keijsers","doi":"10.1109/TNSRE.2026.3655791","DOIUrl":"10.1109/TNSRE.2026.3655791","url":null,"abstract":"Daily life gait performance measures can provide ecologically valid gait characteristics, which are interesting for monitoring individuals with gait impairments. The first step in obtaining these gait characteristics is selecting walking periods from multiple day recordings. We developed and validated an algorithm for walking detection using inertial measurement units (IMU) on Both feet and compared the performance with two others in healthy individuals and those with neurologically impaired gait: 1) using Sacrum accelerometer data (Iluz et al. 2014); 2) using Single foot gyroscopic data (Ullrich et al. 2020). We also assessed which activities reduce the algorithms’ specificity for walking detection. The Both feet algorithm consisted of three stages, 1) selecting time periods potentially containing walking; 2) excluding periods not containing walking; 3) checking the selected periods for minimal walking bout requirements. For validation, 32 participants (12 healthy and 20 with neurologically impaired gait) performed 20-30 minutes of daily life activities, while wearing IMUs on both feet and the sacrum. Using labelled video recordings as reference, we calculated each algorithm’s specificity, sensitivity and accuracy for walking detection. Both feet outperformed the other algorithms on specificity (96.6% versus 92.1% and 72.1% for the Single foot en Sacrum respectively). Stair climbing was misclassified as walking most often by all algorithms. Sacrum outperformed the others on sensitivity (99.5%), but had low specificity and accuracy. The high specificity of the Both feet algorithm makes it suitable when spatiotemporal gait characteristics are of interest, and is applicable in populations with mild neurological conditions affecting gait.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"931-940"},"PeriodicalIF":5.2,"publicationDate":"2026-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11358952","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146003336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-15DOI: 10.1109/TNSRE.2026.3654843
Adriana Cancrini;Bruno Borghi;Naveed Reza Aghamohammadi;Arturo Ramirez;James L. Patton
Characterizing each person’s sensorimotor profile is crucial for designing precise and personalized motor rehabilitation therapies. Building on our previous work in system identification of human motor control dynamics, we now extend our parameter recovery technique developed in synthetic models to a real-world human experiment. This twin-based digital method actively guides the experimental design by selecting the most informative perturbations and movement conditions to most accurately identify (recover) sensory feedback gains. We applied this framework to 10 neurotypical participants, analyzing their performance during arm planar reaching movements. By combining the optimized experimental design with this forward–inverse modeling pipeline, we estimated individual sensory feedback gains. These gains were then used to simulate movement trajectories, achieving a movement prediction accuracy of 85% compared to withheld trajectories performed by the same subjects. These results validate the ability of our mathematical model to capture and explain individual sensorimotor dynamics through the identification of subject-specific feedback gains. This approach offers a promising tool for gaining insights into the roles of different sensory channels and identifying the most informative data required for efficient assessment.
{"title":"Defining Experimental Design for Human Motor Control Identification: A Novel Framework","authors":"Adriana Cancrini;Bruno Borghi;Naveed Reza Aghamohammadi;Arturo Ramirez;James L. Patton","doi":"10.1109/TNSRE.2026.3654843","DOIUrl":"10.1109/TNSRE.2026.3654843","url":null,"abstract":"Characterizing each person’s sensorimotor profile is crucial for designing precise and personalized motor rehabilitation therapies. Building on our previous work in system identification of human motor control dynamics, we now extend our parameter recovery technique developed in synthetic models to a real-world human experiment. This twin-based digital method actively guides the experimental design by selecting the most informative perturbations and movement conditions to most accurately identify (recover) sensory feedback gains. We applied this framework to 10 neurotypical participants, analyzing their performance during arm planar reaching movements. By combining the optimized experimental design with this forward–inverse modeling pipeline, we estimated individual sensory feedback gains. These gains were then used to simulate movement trajectories, achieving a movement prediction accuracy of 85% compared to withheld trajectories performed by the same subjects. These results validate the ability of our mathematical model to capture and explain individual sensorimotor dynamics through the identification of subject-specific feedback gains. This approach offers a promising tool for gaining insights into the roles of different sensory channels and identifying the most informative data required for efficient assessment.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"721-731"},"PeriodicalIF":5.2,"publicationDate":"2026-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11355670","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145984652","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-14DOI: 10.1109/TNSRE.2026.3654400
Kimji N. Pellano;Inga Strümke;Daniel Groos;Lars Adde;Pål Haugen;Espen Alexander F. Ihlen
Cerebral Palsy (CP) is a prevalent motor disability in children, for which early detection can significantly improve treatment outcomes. While skeleton-based Graph Convolutional Network (GCN) models have shown promise in automatically predicting CP risk from infant videos, their “black-box” nature raises concerns about clinical explainability. To address this, we introduce a perturbation framework tailored for infant movement features and use it to compare two explainable AI (XAI) methods: Class Activation Mapping (CAM) and Gradient-weighted Class Activation Mapping (Grad-CAM). First, we identify significant and non-significant body keypoints in very low and very high risk infant video snippets based on the XAI attribution scores. We then conduct targeted velocity and angular perturbations, both individually and in combination, on these keypoints to assess how the GCN model’s risk predictions change. Our results indicate that velocity-driven features of the arms, hips, and legs appear to have a dominant influence on CP risk predictions, while angular perturbations have a more modest impact. Furthermore, CAM and Grad-CAM show partial convergence in their explanations for both low and high CP risk groups. Our findings demonstrate the use of XAI-driven movement analysis for early CP prediction, and offer insights into potential movement-based biomarker discovery that warrant further clinical validation.
{"title":"Toward Biomarker Discovery for Early Cerebral Palsy Detection: Evaluating Explanations Through Kinematic Perturbations","authors":"Kimji N. Pellano;Inga Strümke;Daniel Groos;Lars Adde;Pål Haugen;Espen Alexander F. Ihlen","doi":"10.1109/TNSRE.2026.3654400","DOIUrl":"10.1109/TNSRE.2026.3654400","url":null,"abstract":"Cerebral Palsy (CP) is a prevalent motor disability in children, for which early detection can significantly improve treatment outcomes. While skeleton-based Graph Convolutional Network (GCN) models have shown promise in automatically predicting CP risk from infant videos, their “black-box” nature raises concerns about clinical explainability. To address this, we introduce a perturbation framework tailored for infant movement features and use it to compare two explainable AI (XAI) methods: Class Activation Mapping (CAM) and Gradient-weighted Class Activation Mapping (Grad-CAM). First, we identify significant and non-significant body keypoints in very low and very high risk infant video snippets based on the XAI attribution scores. We then conduct targeted velocity and angular perturbations, both individually and in combination, on these keypoints to assess how the GCN model’s risk predictions change. Our results indicate that velocity-driven features of the arms, hips, and legs appear to have a dominant influence on CP risk predictions, while angular perturbations have a more modest impact. Furthermore, CAM and Grad-CAM show partial convergence in their explanations for both low and high CP risk groups. Our findings demonstrate the use of XAI-driven movement analysis for early CP prediction, and offer insights into potential movement-based biomarker discovery that warrant further clinical validation.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"750-766"},"PeriodicalIF":5.2,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11352985","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145984698","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In multiple vision-demanding tasks, accurately controlling a prosthetic hand to approach a target object is particularly challenging for amputees, as visual attention diverted by other tasks forces them to rely heavily on peripheral vision. This study aims to initially validate that functionally effective sensory feedback can enhance the control of prosthetic hands during object approach under divided visual attention. To quantify prosthesis users’ ability to approach and manipulate objects using central and peripheral vision in real-life scenarios, we conducted two experimental tasks—APPROACHING and PINCH—under two visual feedback modes: full-vision and partial-vision. During the approaching process, we compared four feedback conditions: no supplementary sensory feedback (PURE), traditional continuous feedback (CONT), evenly distributed discrete feedback (ADIS), and a novel discrete strategy based on Weber’s law (WDIS) proposed in this study. Task performance was evaluated using metrics such as position error, dispersion, task completion time, and pinch failures, while psychological factors were assessed through a questionnaire. Results show that WDIS enabled more accurate and stable object approach, with shorter task completion times, which leads to better subsequent manipulation performance. This also provides participants with enhanced psychological experiences, including reduced workload and increased intuitiveness. WDIS improved prosthetic control and user experience in the simplified laboratory settings, providing a foundation for real-world applications.
{"title":"Discrete Tactile Feedback Based on Weber’s Law Enhances Prosthetic Hand Approaching Performance Under Divided Visual Attention","authors":"Xianwei Meng;Jianjun Meng;Guohong Chai;Xinjun Sheng;Xiangyang Zhu","doi":"10.1109/TNSRE.2026.3653788","DOIUrl":"10.1109/TNSRE.2026.3653788","url":null,"abstract":"In multiple vision-demanding tasks, accurately controlling a prosthetic hand to approach a target object is particularly challenging for amputees, as visual attention diverted by other tasks forces them to rely heavily on peripheral vision. This study aims to initially validate that functionally effective sensory feedback can enhance the control of prosthetic hands during object approach under divided visual attention. To quantify prosthesis users’ ability to approach and manipulate objects using central and peripheral vision in real-life scenarios, we conducted two experimental tasks—APPROACHING and PINCH—under two visual feedback modes: full-vision and partial-vision. During the approaching process, we compared four feedback conditions: no supplementary sensory feedback (PURE), traditional continuous feedback (CONT), evenly distributed discrete feedback (ADIS), and a novel discrete strategy based on Weber’s law (WDIS) proposed in this study. Task performance was evaluated using metrics such as position error, dispersion, task completion time, and pinch failures, while psychological factors were assessed through a questionnaire. Results show that WDIS enabled more accurate and stable object approach, with shorter task completion times, which leads to better subsequent manipulation performance. This also provides participants with enhanced psychological experiences, including reduced workload and increased intuitiveness. WDIS improved prosthetic control and user experience in the simplified laboratory settings, providing a foundation for real-world applications.","PeriodicalId":13419,"journal":{"name":"IEEE Transactions on Neural Systems and Rehabilitation Engineering","volume":"34 ","pages":"674-685"},"PeriodicalIF":5.2,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11348986","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145966025","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：