*Result*: Real-Time OpenSim via IMUs for Full Body Kinematics During Gait, Sports, Exercise, and Dance Movements.

Title:
Real-Time OpenSim via IMUs for Full Body Kinematics During Gait, Sports, Exercise, and Dance Movements.
Source:
IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society [IEEE Trans Neural Syst Rehabil Eng] 2026; Vol. 34, pp. 650-662.
Publication Type:
Journal Article
Language:
English
Journal Info:
Publisher: IEEE Country of Publication: United States NLM ID: 101097023 Publication Model: Print Cited Medium: Internet ISSN: 1558-0210 (Electronic) Linking ISSN: 15344320 NLM ISO Abbreviation: IEEE Trans Neural Syst Rehabil Eng Subsets: MEDLINE
Imprint Name(s):
Original Publication: Piscataway, NJ : IEEE, c2001-
Entry Date(s):
Date Created: 20260112 Date Completed: 20260122 Latest Revision: 20260123
Update Code:
20260130
DOI:
10.1109/TNSRE.2026.3653477
PMID:
41525551
Database:
MEDLINE

*Further Information*

*Despite the growing demand for healthcare services due to an aging population, patients often avoid traditional rehabilitation centers due to high costs, time constraints, and discomfort experienced in laboratory or hospital settings. Home-based rehabilitation offers a promising alternative, but real-time kinematic monitoring and assessment remain challenging. We thus propose a real-time, wireless, portable approach for computing full-body kinematics through OpenSim. Twenty-two subjects performed walking, running, squatting, boxing, yoga, dance, badminton, stair climbing, and seated extremity exercise movements, while wearing 12 SageMotion inertial measurement units (IMUs). Real-time IMU kinematics were computed at 20 Hz and offline kinematics at 100 Hz and were compared with reference optical motion capture kinematics to determine accuracy. Real-time walking and stair climbing were most accurate, both with median RMSE of 5.3 deg. The most accurate joint angle was lumber rotation with median RMSE of 2.7 deg, and the overall median RMSE for all activities across all joints was 7.4 deg. Overall mean RMSE between real-time and offline IMU estimation was 0.7 deg, and mean latency from IMU data reception at the processing hub to kinematics generation was 31.7 ms. This approach could dramatically improve clinical and remote care by enabling rapid assessment and real-time biofeedback for rehabilitation, with potential to significantly enhance patient assessment and treatment outcomes.*