Abstract:Rehabilitation devices, such as traditional rigid exoskeletons or exosuits, have been widely used to rehabilitate upper limb function post-stroke. In this paper, we have developed an exosuit with four degrees of freedom to enable users to involve more joints in the rehabilitation process. Additionally, a hybrid electroencephalogram-based (EEG-based) control approach has been developed to promote active user engagement and provide more control commands.The hybrid EEG-based control approach includes steady-state visual evoked potential (SSVEP) paradigm and motor imagery (MI) paradigm. Firstly, the rehabilitation movement was selected by SSVEP paradigm, and the multivariate variational mode decomposition (MVMD) and canonical correlation analysis (CCA) method was used for SSVEP EEG recognition; then, the motion intention was obtained by MI paradigm, and the convolutional neural network (CNN) and long short-term memory network (LSTM) were used to build a CNN-LSTM model for MI EEG recognition; finally, the recognition results were translated into control commands of Bowden cables to achieve multi-degree-of-freedom rehabilitation.Experimental results show that the average classification accuracy of the CNN-LSTM model reaches to 90.07% ± 2.23%, and the overall accuracy of the hybrid EEG-based control approach reaches to 85.26% ± 1.95%. The twelve subjects involved in the usability assessment demonstrated an average system usability scale (SUS) score of 81.25 ± 5.82. Additionally, four participants who underwent a 35-day rehabilitation training demonstrated an average 10.33% increase in range of motion (ROM) across 4 joints, along with a 11.35% increase in the average electromyography (EMG) amplitude of the primary muscle involved.The exosuit demonstrates good accuracy in control, exhibits favorable usability, and shows certain efficacy in multi-joint rehabilitation. Our study has taken into account the neuroplastic principles, aiming to achieve active user engagement while introducing additional degrees of freedom, offering novel ideas and methods for potential brain-computer interface (BCI)-based rehabilitation strategies and hardware development.Clinical impact: Our study presents an exosuit with four degrees of freedom for stroke rehabilitation, enabling multi-joint movement and improved motor recovery. The hybrid EEG-based control approach enhances active user engagement, offering a promising strategy for more effective and user-driven rehabilitation, potentially improving clinical outcomes.Clinical and Translational Impact Statement: By developing an exosuit and a hybrid EEG-based control approach, this study enhances stroke rehabilitation through better user engagement and multi-joint capabilities. These innovations consider neuroplasticity principles, integrating rehabilitation theory with rehabilitation device.

A Real-Time CNN-BiLSTM-based Classifier for Patient-Centered AR-SSVEP Active Rehabilitation Exoskeleton System

Research on the Control Method of an Upper-limb Rehabilitation Exoskeleton Based on Classification of Motor Imagery EEG

SSVEP based brain-computer interface controlled functional electrical stimulation system for knee joint movement

A portable SSVEP-BCI system for rehabilitation exoskeleton in augmented reality environment

Lower Limb Rehabilitation Exoskeleton Control Based on SSVEP-BCI

An Asynchronous SSVEP-BCI for Intelligent Rehabilitation Exoskeleton Based on Alpha Rhythm-Detected CCA

Multimodal brain-controlled system for rehabilitation training: Combining asynchronous online brain-computer interface and exoskeleton

Motor Imagery Classification Based on Deep Convolutional Neural Network and Its Application in Exoskeleton Controlled by EEG

An Upper-Limb Rehabilitation Exoskeleton System Controlled by MI Recognition Model With Deep Emphasized Informative Features in a VR Scene

Application of a Brain–Computer Interface System with Visual and Motor Feedback in Limb and Brain Functional Rehabilitation after Stroke: Case Report

Design of Exoskeleton Brain-Like Intelligent Trajectory Prediction Model Based on Echo State Network

A 4-DOF Exosuit Using a Hybrid EEG-Based Control Approach for Upper-Limb Rehabilitation

Online Brain-Computer Interface Controlling Robotic Exoskeleton for Gait Rehabilitation

A Synchronous and Closed-Loop Architecture of Bci-Based Rehabilitation System for Stroke with Robot and Virtual Reality

An EEG-driven Lower Limb Rehabilitation Training System for Active and Passive Co-stimulation.

SSVEP Based Brain-Computer Interface Controlled Functional Electrical Stimulation System for Upper Extremity Rehabilitation

Coordinative Motion-Based Bilateral Rehabilitation Training System with Exoskeleton and Haptic Devices for Biomedical Application

Development of an EMG-Controlled Knee Exoskeleton to Assist Home Rehabilitation in a Game Context.

A Hierarchical Intelligent Rehabilitation Robotic System Based on MI-EEG

An Adaptive Hybrid Brain–Computer Interface for Hand Function Rehabilitation of Stroke Patients

BCI Integrated with VR for Rehabilitation