Mei-Hsiang Chen
Chung Shan Medical University, Taiwan
Abstract Title: Feasibility of AI?Driven Personalized Precise Exercise Prescriptions for Community-Dwelling Older Adults
Biography: Professor Chen Mei?Hsiang is an expert in occupational therapy and rehabilitation technology, specializing in upper-limb stroke rehabilitation, virtual reality, and assistive-device design. Research interests in occupational therapy for physical dysfunction (stroke rehabilitation), vocational rehabilitation, and psychometric properties/outcome measurements. She has published numerous SCI/SSCI papers covering outcome-measure validation, clinical trials, and machine-learning applications, serving as corresponding or first author on multiple international journal and conference articles. Her research integrates technology with clinical practice to enhance rehabilitation outcomes and advance occupational therapy education, focusing on evidence-based interventions and practical solutions that improve patient function and professional training.
Research Interest: Exercise serves as a critical therapeutic intervention for enhancing functional capacity in middle-aged and older populations. Comprehensive physiological profiling is essential to quantify an individual’s ability to adapt to physical stressors and activities of daily living (ADLs). This study leveraged AI (Open AI GPT-4.0) to design individualized, precision exercise prescriptions integrated with Nintendo Switch Ring Fit Adventure for community-dwelling older adults. Participants underwent six supervised training sessions based on AI generated protocols to evaluate the feasibility and short-term clinical efficacy of this digital intervention. Preliminary analysis (17 older Adults) demonstrated statistically significant improvements in upper limb muscle strength (biceps arm curl in 30 seconds; mean 2.06, SD 2.13, p = 0.001), Lower limb muscle strength (stand up and sit down in 30 seconds; mean 3.77, SD 1.92, p = 0.000), static balance ability (standing on one leg with eyes open; mean 33.03, SD 39.02, p = 0.003) and Lower limb softness (sitting forward bend in chair; mean 2.88, SD 2.80, p = 0.001). Although there was no significant difference in the progress index of other functional indicators, it may be due to the small sample size and short intervention period. The results of this study indicate that exercise exercises with AI precise exercise prescription can effectively improve muscle strength, balance and softness results. Given the highly heterogeneous health status of the elderly population, precision motion oncology and rehabilitation models are critical to optimizing treatment outcomes and long-term healthspan. Future research will increase the number of samples and training times to further confirm the strength of the effect.