Estimating Knee Joint Contact Forces During Daily Activities Using a Five-IMU Setup

Alessandro Castellaz, Frank J. Wouda, Bert-Jan F. van Beijnum

Confirmation: I have read and agree with the IEEE BSN 2025 conference submission's policy on behalf of myself and my co-authors.

Keywords: Knee joint contact force, inertial measurement units, musculoskeletal modeling, ground reaction forces, biomechanics

Abstract: The estimation of internal joint loading is critical for assessing musculoskeletal health in daily life. This study proposes a simplified method for estimating knee joint contact forces (JCF) using only five inertial measurement units (IMUs) placed on the pelvis, upper legs, and lower legs. Data were collected from three healthy adults performing activities of daily living (ADL) in a home-like environment. Kinematics and ground reaction forces (GRF) were estimated using segment orientations and accelerations, and the virtual pivot point (VPP) concept. Joint reaction forces were then computed using a musculoskeletal model in OpenSim through static optimization and joint reaction analysis. Results showed strong correlations, larger than 0.79, with a mean Root Mean squared Error (RMSE) of 1.06 body weight (BW), that is about 15% relative to the maximum reference range, compared to a reference full-body system. Some error peaks occurred before toe-off, likely due to kinematic inaccuracies. These findings support the feasibility of accurate JCF estimation with minimal instrumentation, enabling real-world monitoring and potential clinical applications. Future work will address calibration improvements to reduce localized estimation errors.

Track: 2. Sensors and systems for digital health, wellness, and athletics

NominateReviewer:

Alessandro Castellaz, [e-mail]
Frank Wouda [e-mail]

Submission Number: 16