AI-driven cognitive telerehabilitation for stroke: a randomized controlled trial

Area of Science:

• Neuroscience
• Rehabilitation Medicine
• Artificial Intelligence

Background:

• Cognitive impairment is a frequent post-stroke complication, necessitating effective rehabilitation.
• Telerehabilitation presents a viable alternative to in-person therapy, but often lacks personalization and engagement monitoring.
• This study addresses the need for advanced telerehabilitation solutions for stroke patients.

Purpose of the Study:

• To evaluate the clinical effectiveness of a self-guided AI-driven cognitive telerehabilitation system.
• To compare the AI-driven system against traditional therapist-supervised rehabilitation in subacute stroke patients.
• To assess the non-inferiority of AI-driven telerehabilitation.

Main Methods:

• A multicenter, randomized controlled non-inferiority trial involving 63 subacute stroke patients with cognitive impairment.
• Participants were assigned to either self-guided AI-driven or therapist-supervised telerehabilitation groups (24 sessions over 6 weeks).
• Primary outcomes included cognitive function tests (K-MMSE2, Trail Making Tests, Digit Span Tests); non-inferiority was tested using K-MMSE2.

Main Results:

• Fifty-five participants completed the trial, with both groups showing significant cognitive improvements.
• No statistically significant differences in cognitive function were observed between the self-guided AI and therapist-supervised groups.
• The AI-driven telerehabilitation was confirmed as non-inferior to therapist-supervised rehabilitation, with high user satisfaction and no adverse events.

Conclusions:

• Self-guided AI-driven telerehabilitation is clinically effective for improving cognitive function in subacute stroke patients.
• The AI system provides comparable cognitive benefits to traditional therapist-supervised rehabilitation.
• AI-powered telerehabilitation offers a promising, accessible, and effective solution for post-stroke cognitive recovery.