This study evaluated a personalized, closed-loop real-time reinforcement feedback strategy for human-machine interface (HMI) control across 5 experiments involving 106 participants (including chronic stroke patients) and two control interfaces, examining effects on force control, retention, and underlying mechanisms via information-theoretic analyses.
Fewer than 20 reinforcement trials produced immediate improvements in force control and lasting retention gains in healthy participants; effects were strongest under limited visual and/or somatosensory feedback. In chronic stroke patients, real-time reinforcement improved online force control under limited visual feedback, though short training did not yield retention gains.
Short training duration may be insufficient to produce retention gains in clinical populations (stroke patients showed no retention benefit). The study involved only two control interfaces, which may limit generalizability to other HMI systems. Stroke sample characteristics and longer-term follow-up outcomes are not reported.
Clinicians developing rehabilitation or assistive technology protocols should consider real-time reinforcement feedback, especially when sensory feedback is sparse — even brief exposure (<20 trials) can quickly improve motor control. This approach shows early promise for stroke rehabilitation but may need longer training to achieve durable gains in patients.
Explore related topics
