The influence of scaling factors and markers’ weighting in inverse kinematics for human motion analysis

¹ Department of Engineering Science and Biomedical Engineering, The University of Auckland, New Zealand
² Poltekkes Kemenkes Jakarta I, Ministry of Health of Republic Indonesia, Indonesia
³ Auckland Bioengineering Institute, New Zealand

^* Corresponding author email: sfau672@aucklanduni.ac.nz

Abstract

Motion analysis can produce variability due to inconsistency in anatomical markers placements, which can lead to misdiagnosis and affect treatment outcomes. This study examined the impact of scaling and marker weighting on repeatability when acquired inverse kinematic (IK) assessment. OpenSim was used to inversely transform the motion capture outputs to assess joint angles, hip, knee, and ankle. One young-healthy participant was included, assessed by five raters. Uniform body segments parameters and different weighting schemes (equal, 10, and 100) for targeted virtual markers were set before static and dynamic data examination. Joint angles were then quantified accordingly, while the statistical analysis was used to test variability among raters. Significant differences were observed between all joint angles with equal-weighted and weighted models, particularly for the hip and knee joints. Root Mean Square Error (RMSE) values indicated notable variability in knee joint angles with a shank weight of 100 (20.23°). Hip joint angles also showed high variability across all conditions, while ankle joint angles had lower overall variability but showed moderate increment throughout gait cycle. Although all raters demonstrated strong agreement, the variability introduced by different weighting schemes highlights the need for careful markers’ weight selection to minimize error. This study demonstrates that scaling and marker weighting in OpenSim can reduce rater-dependent variability, thereby enhancing the consistency of motion capture analysis.