การพัฒนาระบบประเมินสมดุลท่านั่งเชิงปริมาณโดยอาศัยเซนเซอร์และการวิเคราะห์ท่าทางด้วยคอมพิวเตอร์วิทัศน์

Abstract

This study aimed to develop and evaluate the feasibility of a prototype system for quantitative sitting balance assessment by integrating load cell data for analyzing weight distribution during sitting and image processing using MediaPipe for real-time posture analysis from video recordings. The collected data were analyzed based on the framework of the Sitting Balance Scale (SBS), a standard clinical tool for sitting balance assessment, to design quantitative indicators corresponding to selected SBS assessment items. Laboratory-based experiments were conducted as a proof-of-concept and feasibility study under controlled conditions to evaluate the system’s ability to detect trunk inclination, weight-shifting patterns, and reaching in different directions while sitting. Preliminary results demonstrated that the system could accurately measure balance-related variables in accordance with actual postural behaviors and convert sitting balance performance into continuous and reproducible numerical data. These findings highlight the potential of the proposed system as an objective tool for quantitative sitting balance assessment, which can complement traditional SBS-based evaluations and facilitate long-term monitoring of changes in postural control and balance performance.