Aug 2026· International Conference on Advanced Mechatronic Systems· pp. 103-108· 0 citations· 36 references
Abstract
Electroencephalography-based brain-computer interfaces (EEG-based BCIs) provide a non-invasive pathway for incorporating voluntary neural activity into lower-limb rehabilitation robots, exoskeletons, robotic orthoses, and gaittraining systems. This focused review examines the MI-based brain-robot rehabilitation loop, including lower-limb intention decoding, high-level command generation, robot or gait-device interaction, and closed-loop feedback. Evidence is interpreted across four categories: offline lower-limb MI decoding, online BCI demonstrations, lower-limb device integration, and patient-oriented or clinical evaluation. Representative studies support the technical feasibility of decoding lower-limb motor imagery (MI) and using selected BCI outputs in virtual-reality, exoskeleton, and treadmill systems. However, the evidence remains dominated by offline analyses and small proof-of-concept studies, with limited standardized clinical outcomes. Hybrid sensing and multimodal feedback have been explored as complementary strategies, but their value in physical lower-limb rehabilitation requires direct online and patient-oriented validation. The review therefore distinguishes transferable decoding advances from direct rehabilitation evidence and identifies priorities for safe and clinically meaningful system development.
The neurophysiological basis of EEG-BCI and three major rehabilitation paradigms are outlined, including motor imagery with physical feedback, motor imagery with virtual/multisensory feedback, and the steady-state visual evoked potential (SSVEP)-driven paradigm.
Wang Peng, Yang Yang, Juehan Wang et al.· Topics in Stroke Rehabilitat...· 0 citations
Population aging and the rising prevalence of motor disorders are driving demand for assistive lower-limb robotic systems capable of decoding user intention rather than merely providing mechanical support. We present CM-FuseNet, an attention-augmented hybrid Brain–Computer–Muscle Interface (BCMI) that simultaneously fuses cortical concentration indices extracted from electroencephalography (EEG) and lower-limb intention patterns derived from electromyography (EMG) to adaptively control a 4-DOF assistive lower-limb exoskeleton. To eliminate the burden of human-subject ethics review and to ensure reproducibility of the proposed methodology, all validation is performed exclusively on (i) permissively licensed open-access biomedical datasets, (ii) high-fidelity OpenSim 4.5 and MuJoCo 3.1 musculoskeletal–exoskeleton co-simulation, and (iii) limited self-experimentation by the corresponding author with non-invasive consumer-grade devices. Three components are introduced: (i) a log-tanh normalized concentration index CI in (0, 1) derived from the (PSMR+PMid−Beta)/PTheta ratio; (ii) a bidirectional Cross-Modal Transformer (CMT) with eight-head self- and cross-attention; and (iii) a Soft Actor-Critic (SAC) reinforcement-learning controller that adaptively tunes four servo PID gains using a concentration-weighted state. Experiments on the PhysioNet EEGMMIDB, Ninapro DB2/DB7, HuMoD and WAY-EEG-GAL datasets (combining N = 162 trial sessions, 47,520 windows, and five-fold cross-validation) yield a gait-phase classification accuracy of 96.84 ± 1.18%, torque-tracking RMSE of 0.072 ± 0.008 N·m, information transfer rate of 38.6 bits/min, end-to-end latency of 9.4 ms, and a 27.4% reduction in simulated metabolic cost over an EMG-only PID baseline (one-way ANOVA: F(4, 75) = 47.83, p < 0.001; Tukey HSD: p < 0.01 against all baselines). Under high cognitive load, CM-FuseNet preserves accuracy with only a 4.63 percentage-point degradation versus 13.22 percentage points for the EMG-only baseline.
Yongseong Park, D. Shin, Hun-kee Kim· Applied Sciences· 0 citations
MI-BCI training can improve upper limb motor function, particularly for isolated movements and fine motor control, in stroke patients, but the current evidence does not support definitive conclusions regarding its superiority over standardised traditional rehabilitation.
Zhen Yang, Shan Zhang, Du Wang et al.· Brain Impairment· 0 citations
Addressing the demand for portable, real-time brain-computer interface systems in stroke rehabilitation, this chapter completes the physical integration and online experimental validation of a wearable system. The system utilizes a specialized EEG headset with miniaturized acquisition circuits secured via pogo pins, featuring 10 core recording channels strategically positioned over the sensorimotor cortex. During the experimental evaluation phase, the research team recruited 6 healthy subjects and 2 stroke-affected hemiplegic patients to conduct closed-loop experiments based on motor imagery and motor attempts. This study employed Common Space Pattern (CSP) for spatial feature extraction, combined with Linear Discriminant Analysis (LDA) for intention classification. Recognition performance was further enhanced through personalized sub-band optimization techniques. Experimental results demonstrate an average offline recognition rate of 84.91%. In the more challenging online real-time testing, the classification accuracy reached 79.38%. Furthermore, by analyzing spatiotemporal spectra and R² value distributions, the study validated the activation patterns in the brain's sensorimotor areas during motor intention triggering at the neurophysiological level. The findings in this chapter provide critical data support and technical solutions for advancing brain-computer interface technology from laboratory settings to community rehabilitation.
Liying Zhang, Jia-Shan Li, Yifei Wang et al.· ITM Web of Conferences· 0 citations
Electroencephalogram (EEG)-based brain-computer interfaces (BCIs) have been widely explored for detecting and monitoring mental health-related states, with many existing studies focusing on identification and classification. Such approaches are primarily observational and provide limited support for intervention. Closed-loop EEG-based BCIs address this limitation by incorporating real-time feedback to observe neural activity. A key paradigm within this context is neurofeedback, in which users learn to modulate their own brain activity, with the aim of supporting improvements in mental states and in psychological functioning. In this work, we conducted a review of closed-loop EEG-based BCI interventions for mental health published since 2021, guided by the preferred reporting items for systematic reviews and meta-analyses (PRISMA) reporting principles. A structured search was conducted across three databases (Scopus, Web of Science, and PubMed), yielding 1,101 records, of which 25 studies met the inclusion criteria. Advancements and observations were summarized across four categories: application, paradigm design and feedback mechanisms, signal processing and machine learning methods, and performance metrics and outcomes. In addition, this review discusses considerations related to signal processing and machine learning (ML), user interface design, and regulatory mechanisms across BCI interventions. Finally, potential directions for future research are outlined, including multimodal BCIs, domain adaptation, the integration of generative AI for BCI-based therapeutic interventions, and home-based deployment. Overall, current findings support the technical feasibility and emerging therapeutic potential of closed-loop EEG-based neurofeedback and BCI interventions in mental health applications, but the evidence base remains preliminary, as many studies were small pilot or feasibility studies.
Yue Zhang, Kun Qian, Damien Coyle et al.· Frontiers in Neuroscience· 0 citations
Post-stroke wrist-hand rehabilitation robots can increase training intensity, repetition, and measurement. Still, systems centered mainly on mechanical assistance may not fully address the sensorimotor impairments that constrain functional recovery. This structured narrative review examines how sensorimotor closed-loop control may reshape wrist-hand robotic rehabilitation by linking somatosensory feedback, voluntary motor effort recognition, and home-based translation. The review was informed by a PubMed-based literature pool and organized around three interrelated themes rather than a meta-analytic effect estimate. Current evidence suggests that proprioceptive, tactile, vibrotactile, and electro-vibro-feedback strategies may support sensorimotor integration. At the same time, EMG-, EEG/BCI-, and CMC-informed approaches can help detect residual motor intention or effort and trigger robotic, FES, or NMES assistance. Home-based and telerehabilitation systems also show potential to improve access, training dose, and therapist-supported monitoring. However, the evidence remains heterogeneous, often limited by small samples, pilot designs, mixed devices, and incomplete validation of long-term home use. Future wrist-hand systems should integrate somatosensory feedba ck, voluntary effort recognition, adaptive robotic/FES/NMES assistance, objective neurophysiological and kinematic assessment, home monitoring, and therapist-in-the-loop supervision, while avoiding claims of clinical superiority before stronger controlled evidence is available.
