AI-Guided Tele-Rehabilitation for Children with Congenital Talipes Equinovarus Following the Ponseti Method: A Prospective Cohort Study on Feasibility and Efficacy

Background: Congenital talipes equinovarus (CTEV) is a common congenital musculoskeletal deformity. While the Ponseti method achieves excellent initial correction, maintaining outcomes during the bracing and rehabilitation phase remains challenging due to poor adherence and inconsistent home exercise quality. This study evaluated the feasibility and efficacy of a novel artificial intelligence (AI)-guided tele-rehabilitation system designed to support home-based management following the Ponseti protocol. Methods: A prospective cohort study was conducted with 55 children (aged 3–5 years) post-Ponseti casting who wereassigned to a 12-month AI-guided tele-rehabilitation program (AI-TR, n=28) or conventional home rehabilitation (TR, n=27). The AI system used smartphone-based computer vision to provide real-time feedback on exercise techniquesand monitor brace adherence. The primary outcomes were ankle dorsiflexion and heel valgus angles. Secondary outcomes included Laaveg-Ponseti functional scores, brace compliance, parental anxiety (Generalized Anxiety Disorder 7-item，GAD-7), parental knowledge, child sleep quality (Pittsburgh Sleep Quality Index，PSQI), adverse events, and satisfaction. Assessments were performed at baseline （1 month post-discharge）, 6 months （6 months post-discharge）, and 12 months （12 months post-discharge）. Trial Registration: ChiCTR2600117423. Retrospectively registered on 23/01/2026 . Results: Compared with the TR group, the AI-TR group showed significantly greater improvement in ankle dorsiflexion (15.71° vs. 10.56°, p < 0.001 , Cohen’s d=2.21 ) and heel valgus (20.00° vs. 12.60°, p < 0.001 , d =2.47). Brace compliance was markedly higher in the AI-TR group (92.9% vs. 44.4%, p < 0.001 ). The AI-TR group also achieved superior Laaveg-Ponseti scores, greater reductions in parental anxiety, greaterparental knowledge, and better child sleep quality (all p <0.05). Parental satisfaction with the AI system was high, with 93% willing to recommend it. A lower relapse rate was observed in the AI-TR group (0% vs. 7.4%). Conclusions: An AI-guided tele-rehabilitation system is feasible and effective for improving adherence, functional outcomes, and family-centred measures during the maintenance phase of CTEV rehabilitation. This digital approach offers a scalable strategy to enhance long-term outcomes and the accessibility of clubfoot management.