Technology-Based Complex Motor Tasks Assessment: A 6-DOF Inertial-Based System Versus a Gold-Standard Optoelectronic-Based One

• Published in: IEEE sensors journal Vol. 21; no. 2; pp. 1616 - 1624
• Main Authors: Saggio, Giovanni, Tombolini, Francesca, Ruggiero, Antonio
• Format: Journal Article
• Language: English
• Published: New York IEEE 15.01.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Calibration; Cameras; gait analysis; Gait recognition; Hip; Human motion; IMU; Inertial measurement unit; Inertial sensing devices; Joints (anatomy); Knee; Legged locomotion; optoelectronic analysis; Optoelectronics; Parameters; Pelvis; Reliability; Sensors; Systems analysis; Task complexity; Walking; wearable system
• ISSN: 1530-437X; 1558-1748
• DOI: 10.1109/JSEN.2020.3016642

Currently, the gold-standard method of assessing human motion is by means of optoelectronic analysis systems. However, such systems have some drawbacks (time-consuming procedure, specialized room, expensive,..) and therefore other analysis systems are gaining in importance. Here, we report a novel inertial-sensor based system (Movit System G1, by Captiks) with an innovative calibration, testing its strengths and weaknesses when compared to an optoelectronic gold standard one (Vicon, by Oxford Metrics). In particular, the validation was executed with different subjects performing different motor exercises: walking (Test 1) and joint movements (Test 2). Gathered data from measurements were anlayzed to evidence accuracy and reliability of the inertial-sensor based system, and the statistical agreement between the two measuring approaches. Results demonstrated a very good measurement accuracy of the inertial-based system for hip, knee and ankle's ROMs in the sagittal plane during walking (RMSE <2.66°, PCC >0.97), and for joint movements in all planes (RMSE <3.46°, PCC >0.94). In addition, the two systems performed with a good agreement (the percentage errors of spatio-temporal parameters were lower than 5%, except for double support which was equal to 8.6%). The reliability was proved for the most gait parameters and joints.