Active rehabilitation technologies for stroke patients: a mini-review

Stroke remains one of the leading causes of long-term disability worldwide, with motor dysfunction being one of the most common and persistent consequences. Traditional rehabilitation often relies heavily on passive therapy, which can be limited by insufficient training intensity, low patient engagement, and weak integration of intention, action, and feedback. In contrast, active rehabilitation emphasizes voluntary patient participation and uses interventions based on intention and enriched with feedback to stimulate recovery through neuroplasticity. This review summarizes the physiological basis of active rehabilitation, focusing on the mechanisms of neuroplasticity, motor intention detection, and feedback intervention. Additionally, it explores how rehabilitation strategies should adapt to the acute, subacute, and chronic phases of post-stroke recovery. However, clinical implementation is still hindered by inter-individual variability, signal instability, heterogeneity of protocols, and the difficulty of determining the optimal timing and dosage of intervention. Therefore, an approach that takes into account the stage of the disease and the individual characteristics of the patient is necessary to develop effective rehabilitation systems. In general, active rehabilitation represents a promising link between engineering innovations and clinical neurorehabilitation, and future development directions include multimodal probing, closed-loop intelligent control, home systems, and digital patient twins.