Robotic rehabilitation and intelligent algorithms improving the performance skills of stroke patients: a scoping
Omid Rustamzadeh1, Seyed Ali Hosseini2, Rastegar Rahmani Tanha3
1Student Research Committee, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran; Department of Occupational Therapy, School of Rehabilitation Sciences, University of Social Welfare and Rehabilitation Sciences, Tehran, Iran.
Background:
This scoping review highlights major advances and persisting gaps in robotic and AI-driven rehabilitation for stroke, evaluating their impact on hand strength, dexterity, and ROM, and offering clinicians practical, updated guidance.
Methods:
Studies that focused on robotic-assisted technologies (RATs) in upper limb rehabilitation for stroke survivors (2014-2024) were included. Study designs unrelated to stroke, animal studies, and conference abstracts were excluded. Systematic searching in PubMed, Web of Science, Scopus, and Google Scholar employed robotic rehabilitation, AI, hand function, and stroke recovery-related terms. Data extraction encompassed intervention type, duration of treatment, dosage of therapy, outcome measures, cost-effectiveness, and patient satisfaction. Types of robotic rehabilitation: end-effector robots, exoskeletons, soft robotic gloves (SRGs), brain-computer interfaces (BCIs), and AI-enhanced virtual reality (AIVR).
Results:
These devices can augment motion, grip strength, and functional independence, especially in chronic and subacute stroke patients. Therapies are made fine-grained by algorithms to balance challenge and engagement, thus lightning therapists' burdens. Conventional energy sources may offer a more attractive option at shorter timelines and with reasonably predictable availability. Models that can be done at home enhance adherence at that higher level, though usability appears high for most models. Still, challenges with setup and independence for participants remain.
Conclusion:
Robotic rehabilitation has a significant impact on motor function (MF) among stroke patients. Despite this, obstacles such as cost, accessibility, and long-term efficacy need even more research. Therapy dose optimization, adaptive AI integration, and cognitive-emotional outcome assessment are all areas of gaps in robotic rehabilitation that still need to be addressed.
