Toward Designing Motor Imagery-driven Hand Redirection in Virtual Reality for Upper Limb Rehabilitation

Motor-impaired patients often experience fatigue and low self-efficacy during rehabilitation training, which can hinder their compliance and the therapeutic effect. Hand redirection (HR) in virtual reality (VR) amplifies the visually perceived movement of users' virtual hand relative to the real hand, enabling users to tackle seemingly more challenging tasks, thereby boosting their motivation and exhibiting more effort in rehabilitation. However, prior HR techniques amplify users' hand movements without considering their engagement level, while active mental engagement is vital for stimulating neuroplasticity which enhances rehabilitation effects. To fully engage patients during rehabilitation, we proposed a Motor Imagery (MI)-driven HR technique, which only amplifies the virtual hand movement when users are fully engaged based on active brain signals. Our study with stroke patients revealed that, compared with Default-HR and No-HR, the MI-HR better engages patients' attention, while they perform better than No-HR in both HR conditions. We further discuss the design implications and potential applications of the system in broader rehabilitation scenarios.