Veronica Bessone1, Nadja Höschele1, Ansgar Schwirtz1, Wolfgang Seiberl1,2
Using inertial measurement units (IMUs) in monitoring and analysing sport movements has become popular in sports research since it avoids the laboratory limitation. However, the accuracy of modern IMU-systems (hardware combined with software) needs to be validated using gold-standard systems as baseline. In this study, we investigated the feasibility of the aktos-t IMU-system for in-field biomechanical research by comparing its outputs in various tasks (repetitive movements, gait and jumping) undertaken by 14 participants, with those of an optoelectronic system. The results showed that the accuracy of aktos-t varies according to the task performed. The accuracy of pelvis, hip and knee joints ranged between acceptable (root mean squared error (RMSE) < 5°) and tolerable (RMSE < 10°) in gait, while the upper limb joints showed inaccuracy (RMSE > 10°) and imprecision (coefficient of repeatability > 10°) during the repetitive movement test. Jump impact appeared not to influence the IMU outcomes (p > 0.05). The main sources of error could be related to the IMU-alignment during the reference T-pose. Finally, the study provides researchers the means for evaluating the accuracy of aktos-t (hardware, software and biomechanical model) as sufficiently precise for its application in their in-field investigations.
KEYWORDS: Motion capture, IMU, outdoor measurements, in-field feedback, jump
1Department of Biomechanics in Sports, Faculty of Sport and Helath Sciences, Technical University of Munich, Germany
2Institute of Sport Science, Bundeswehr University Munich, Neubiberg, Germany
50 free eprints available here...
https://www.tandfonline.com/eprint/WPD3DEZNDIVKC8RMNK7A/full?target=10.1080/14763141.2019.1671486