Validation of Artificial Intelligence-prescribed Exercise Programs for Improving Upper Crossed Syndrome and Dynamic Knee Valgus

Abstract

Background: The growing integration of artificial intelligence (AI) in healthcare is creating innovative pathways for tailored health solutions. While AI can generate training programs, the validation of the effectiveness of AI-generated exercise programs remains unexplored. Objectives: Therefore, this study aims to investigate the validation of AI-prescribed exercise programs for improving upper crossed syndrome (UCS) and dynamic knee valgus (DKV). Methods: This study involved developing an AI-generated exercise program utilizing the Delphi method. The Delphi process consists of administering a questionnaire within a specific domain, where a panel of experts assesses the program’s suitability. Three methods were used to determine validity: Content Validity Ratio (CVR), Content Validity Index (CVI), and Impact Score (IS). The Fleiss Kappa coefficient (κ) was calculated to assess the degree of agreement (reliability) between the experts’ responses. Data analysis was performed using SPSS version 27 and Microsoft Excel version 2024. Results: The IS indicates that all exercises possess the required level of validity for UCS and DKV. However, according to the CVI and CVR, while the majority of exercises demonstrated acceptable content validity, a small number did not meet the necessary thresholds. Conclusions: The findings suggest that while platforms like ChatGPT-4o can generate generally appropriate material, discrepancies remain in terms of expert consensus with established validity benchmarks. Therefore, AI may support rehabilitation only as an adjunct under professional supervision, rather than as an independent tool.