Scholarlink

<p>This study addresses the common issue of sarcopenia among older adults by developing a personalized exercise therapy system that integrates MediaPipe-based image recognition with serious games. Utilizing non-wearable posture tracking technology, the system captures human key points in real time and calculates joint angles, combining them with interactive game design to enhance training motivation and consistency. It provides three training modules: bicep curls, overhead presses, and single-leg balance exercises. The system automatically adjusts the difficulty level based on the user’s performance. The exercise analysis platform offers feedback on repetition counts, joint angles, and fatigue trends, helping users identify problems early and continuously optimize training effectiveness. Experimental results indicate significant improvements in participants' performance in bicep curl and balance training, particularly in the increased range of upper limb motion, reduced time to complete a single movement, and improved motion scores. These improvements reflect enhanced muscular strength, refined motor control, and improved dynamic stability, which contribute to better maintenance and enhancement of daily functional abilities. Although changes in the average performance time of the overhead press did not reach statistical significance, the overall performance showed steady improvement, suggesting potential benefits in coordination and endurance. In summary, this system demonstrates the potential to improve muscle strength, balance ability, and movement efficiency, offering evidence-based exercise support for promoting healthy aging in older adults.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development and Evaluation of a Serious Game-Based Exercise Therapy System for Sarcopenia Using Pose Recognition Technology

  • 期刊论文

  • 发表日期：

    2025年8月1日

查看全文
  • Lun-Ping Hung,
  • Shih-Yang Yang,
  • Zong-Jie Wu,
  • Wei-Tong Lin,
  • Mei-Yu Wu

摘要

This study addresses the common issue of sarcopenia among older adults by developing a personalized exercise therapy system that integrates MediaPipe-based image recognition with serious games. Utilizing non-wearable posture tracking technology, the system captures human key points in real time and calculates joint angles, combining them with interactive game design to enhance training motivation and consistency. It provides three training modules: bicep curls, overhead presses, and single-leg balance exercises. The system automatically adjusts the difficulty level based on the user’s performance. The exercise analysis platform offers feedback on repetition counts, joint angles, and fatigue trends, helping users identify problems early and continuously optimize training effectiveness. Experimental results indicate significant improvements in participants' performance in bicep curl and balance training, particularly in the increased range of upper limb motion, reduced time to complete a single movement, and improved motion scores. These improvements reflect enhanced muscular strength, refined motor control, and improved dynamic stability, which contribute to better maintenance and enhancement of daily functional abilities. Although changes in the average performance time of the overhead press did not reach statistical significance, the overall performance showed steady improvement, suggesting potential benefits in coordination and endurance. In summary, this system demonstrates the potential to improve muscle strength, balance ability, and movement efficiency, offering evidence-based exercise support for promoting healthy aging in older adults.