Journal of Robotics, Automation and Smart Systems

ABSTRACT

Objective

To explore the use of artificial intelligence (AI) and robotic technologies in fine motor rehabilitation of the upper limb after stroke and their implications for occupational therapy.

Methods

A scoping review was conducted following the PRISMA-ScR guidelines. A scoping review was conducted following the PRISMA-ScR guidelines. The first author conducted database searches in PubMed/MEDLINE, CINAHL, IEEE Xplore, and the Cochrane Library in [Month Year], using the strategy detailed in Appendix A. The search covered studies published from January 2010 to December 2025. Reference lists of included studies and relevant reviews were hand-searched. All search results were exported, deduplicated, and screened by [Author initials]. The final evidence map was verified against accessible full-text records on June 26, 2026.The studies included adults post-stroke and focused on AI-enabled, brain-computer interface, robotic, sensor-based, or digitally adaptive interventions aimed at improving hand, finger, grasp, release, pinch, or dexterity. Data were collected on population, technology type, AI or digital function, intervention context, outcome measures and main clinical findings. Results: Fourteen intervention studies were included in the final evidence map, which were grouped into four areas: motor-intent detection, adaptive robotic assistance, sensor-based feedback and assessment, and home-based or telerehabilitation support. Common outcomes were the Fugl-Meyer Assessment of the Upper Extremity, Action Research Arm Test, Box and Block Test, grip and pinch strength, Motor Activity Log, and the Canadian Occupational Performance Measure. Most studies reported statistically significant changes in impairment or activity-level outcomes; however, the clinical significance of these changes is uncertain, and the evidence is limited by small sample sizes, diverse technologies, varying intervention doses, and limited occupation-based measurements.

Conclusion

Preliminary evidence suggests that AI and robotic technologies may enhance the intensity, feedback, and personalization of upper limb rehabilitation after stroke, though larger controlled trials are needed to confirm these benefits. Current evidence supports their use as supplements to therapist-led interventions rather than as replacements for occupational therapy. Future research should examine usability, cost, safety, and effects on meaningful hand use in daily activities.

Keywords: Artificial Intelligence; Stroke Rehabilitation; Upper Extremity; Fine Motor Skills; Robotics; Brain-Computer Interface; Occupational Therapy