Biomechanical Analysis of Dancing Movements: A Review of Technologies and the Development of an Artificial Intelligence-Based Injury Prevention System

Abstract

This study presents the design and development of DANZALYZE, an AI-based biome chanical analysis system developed to analyse dance movements, provide posture correction feedback, and support injury prevention among dancers. The system integrates computer vision and web technologies to offer a cost effective and accessible alternative to conventional motion-capture systems. DANZALYZE is based on the MediaPipe Pose Landmark model, which evaluates alignment, balance, and movement precision using 33 human body key points extracted from standard two-dimensional video recordings. A Python-based module is used to perform pose estimation and similarity scoring, while the ASP.NET framework supports user interaction, data processing, and visualization. System evaluation was conducted through quantitative accuracy testing and a usability study, indicating stable consistency with professional instructor ratings. User satisfaction results demonstrated an average System Usability Scale score of 86.5, reflecting positive learning impact and excellent usability. The system provides meaningful real-time feedback that increases user awareness of posture safety and movement quality. The study concludes that AI-based pose estimation systems can be introduced into the field of dance to improve performance monitoring, reduce injury risk, and expand access to biomechanical analysis. Future work includes the adoption of three-dimensional motion analysis, cloud computing, and mobile integration to enhance real-time accessibility. Overall, this system illustrates how technology can bridge art and science through data-driven movement analysis in dance.