A novel real-time computer vision and artificial intelligence based hand function rehabilitation program for children with cerebral palsy.

PURPOSE: To evaluate the effectiveness of a real-time computer vision and artificial intelligence(AI) based hand function rehabilitation program in improving upper extremity function and activities of daily living in children with cerebral palsy(CP). METHODS: A randomized controlled trial was conducted in 45 children with CP aged 18 months to 7 years and 5 months. The intervention group participated in an 8-week rehabilitation program (three 15-minute sessions/week); the control group received no intervention. The primary outcome was the Box and Block Test. Secondary outcomes included the Functional Dexterity Test, pinch and grip strength, Quality of Upper Extremity Skill Test, Pediatric Upper Extremity Motor Activity Log-Revised, and Pediatric Evaluation of Disability Inventory. RESULTS: Significant time-by-group interaction effects were observed for BBT and palmar pinch strength in both hands (p < 0.05), and for FDT and tip pinch strength in non-dominant hand (p < 0.05). Greater improvements in non-dominant hand BBT scores were associated with higher baseline upper extremity function. Mean compliance was 91.9%, and caregivers reported high satisfaction and less burden. CONCLUSIONS: The real-time computer vision and AI based rehabilitation program, offering engaging hand-motion training with enhanced motivation, was improved upper extremity function in children with CP, supporting its feasibility for home-based therapy.