Authors: (including presenting author): :
Poon CKT (1), Leung AKY (1), Chan ASM (1), Yuen KMW (1)
Affiliation: :
(1) Occupational Therapy Department, Alice Ho Miu Ling Nethersole Hospital
Keyword 1: :
Artificial intelligence
Keyword 2: :
Occupational therapy
Keyword 3: :
Fine motor training
Keyword 4: :
Special Educational Needs
Introduction: :
The increasing prevalence and earlier identification of children with special educational needs (SEN) has led to heightened demand for intensive, individualized fine motor (FM) training. This growing demand has placed considerable pressure on hospital-based occupational therapists, resulting in heavier administrative workloads and reduced opportunities for clinical mentorship. Coordinated with the HA IT Innovation Office, this study piloted an artificial intelligence (AI)-assisted occupational therapy (OT) program to streamline documentation processes and provide data-driven, personalized activity recommendations to enhance the efficiency of treatment planning and clinical education.
Objectives: :
This pilot study sought to: 1) assess the feasibility of integrating AI into OT practice; 2) examine its impact on treatment planning and documentation efficiency; 3) evaluate its utility in facilitating mentorship; and 4) measure program acceptability among parents and therapists.
Methodology: :
A single-arm pilot study was conducted with a convenience sample of 40 children aged 4–6 years referred for developmental training. Participants received 6 AI-assisted FM training sessions. Assessment data were recorded on an electronic platform developed by the Hong Kong Applied Science and Technology Research Institute (ASTRI). The platform provided data storage, AI-generated activity recommendations, session video recording for mentorship purposes, and automated OT progress notes generation. Satisfaction questionnaires were administered to parents and therapists post intervention. Pre-and-post FM performance was evaluated using subtests from the Bruininks-Oseretsky Test of Motor Proficiency, Second Edition.
Result & Outcome: :
Findings demonstrated high acceptability of the AI-assisted system. Parents (n=30) reported strong satisfaction (Mean=4.63/5) and willingness to participate in future AI-integrated therapy programs (Mean=4.37/5). Therapists (n=3) highlighted the system’s utility in enhancing documentation efficiency and promoting scoring standardization.
Pre- and post-intervention assessments revealed statistically significant improvements in FM precision (p=.020) and FM integration (p=.033). Therapists’ feedback identified areas for further development, including larger datasets to improve recommendation accuracy, enhanced modelling of clinical reasoning, and a more intuitive user interface. This pilot study supports the feasibility of incorporating AI-assisted models into OT interventions for children with SEN. The program demonstrated potential benefits in improving treatment efficiency, documentation, and mentorship, while yielding positive outcomes in fine motor performance. Future research with larger sample sizes and controlled trials is recommended to further evaluate treatment effectiveness and explore long-term outcomes.
