|
|
An overview of the scientific validation of the Wii Balance Board and the Kinect as clinically relevant measurement devices Host Publication: Symposium on Technology-Supported Rehabilitation Authors: B. Jansen, B. Bonnechère, L. Omelina, V. Sholukha, M. Rooze and S. Van Sint Jan Publication Date: Nov. 2014
Abstract: An overview of the scientific validation of the Wii Balance Board and the Kinect as clinically relevant measurement devices.
Jansen B.2,3, Bonnechère B.1*, Omelina L.2,3, Sholukha V.1,4, Rooze M.1,
Van Sint Jan S.1
1 Laboratory of Anatomy, Biomechanics and Organogenesis (LABO), Université Libre de Bruxelles, CP 610, Lennik Street 808, 1070 Brussels, Belgium
2 Department of Electronics and Informatics - ETRO, Vrije Universiteit Brussel, Pleinlaan 2, B Brussels, Belgium
3 iMinds, Dept. Medical Information Technologies (MIT), Gaston Crommenlaan 8 (box 102), B Ghent, Belgium
4 Department of Applied Mathematics, State Polytechnical University (SPbSPU), Saint Petersburg, Russia
The booming field of exergaming in rehabilitation is benefiting from the wide availability of low-cost motion sensors, like the Kinect camera and the Wii balance board. Both commercially available mainstream games as well as games developed for specific rehabilitation exercises are being introduced in the daily practice.
It remains however an important question how "correct" the skeleton obtained from the Kinect is and how "correct" the center of pressure information from the balance board is. This question is of high importance for the construction of high quality rehabilitation games, but also for the scientific research around it:
(1) Knowledge on the accuracy/quality of the input devices can result in better game design, for instance by avoiding exercises, poses ... that cannot be registered correctly.
(2) Data obtained from the gaming hardware could be considered as a low-cost low-quality motion analysis (Kinect) or low-cost low-quality posturograhy analysis (balance board). Importantly, this allows to quantify the patient's evolution of the course of rehabilitation on a regular basis, without being referred to highly specialized motion assessment laboraties.
With respect to the scientific validation of markerless motion capture solutions as the Kinect for motion analysis, the presentation will focus on the results of four different studies, comparing the MLS and MBS with respect to segment lengths and range of motion, during several clinically relevant tasks.
Segment lengths were computed for both systems and compared in static position (N=48). Results mainly show that there were significant differences between both systems, but that these differences were systematic and can hence be removed by means of regression equations. ICC between both systems was found to be high for most segments, except for foot and trunk [1,2].
For shoulder abduction, elbow flexion, hip abduction and knee flexion, range of motion was compared between both systems (N=48)[3]. For shoulder abduction excellent agreement was observed (R2 = 0,98) for the other 3 exercises poor to no agreement was observed. However, ICC between test and retest for the 4 exercises are high and very similar for both systems.
Finally, a fourth study focused on the analysis of deep squatting (N=27) and found very poor agreement between both systems with respect to segment length, but good agreement for the three most important joint angles during the task (shoulder R2=0,87, hip R2=0,88 and knee R2=0,77) [4].
With respect to the scientific evaluation of the WII balance board, a debate is currently going on in the Gait and Posture journal (april 2014), resulting in almost dogmatic positions. The presentation will try to summarize the currently existing validation studies for motion assessment and will discuss their importance both for clinical assessment as well as for exergames. For the latter, results from the literature will be complemented with our own (unpublished) findings in order to show that at least for dynamic exercises, the balance board is reliable and repeatable, not only for COP measurements, but also for a range of parameters computed in posturography.
[1] Bonnechère, B., Jansen, B., Salvia, P., Bouzahouene, H., Sholukha, V., Cornelis, J., ... Van Sint Jan, S. (2014). Determination of the precision and accuracy of morphological measurements using the KinectTM sensor: comparison with standard stereophotogrammetry. Ergonomics, 57(4), 622ᆳ. doi:10.1080/00140139.2014.884246
[2] Bonnechère, B., Sholukha, V., Jansen, B., Omelina, L., Rooze, M., & Van Sint Jan, S. (2014). Determination of repeatability of kinect sensor. Telemedicine Journal and E-Health?: The Official Journal of the American Telemedicine Association, 20(5), 451Dž. doi:10.1089/tmj.2013.0247
[3] Bonnechère, B., Jansen, B., Salvia, P., Bouzahouene, H., Omelina, L., Moiseev, F., ... Van Sint Jan, S. (2014). Validity and reliability of the Kinect within functional assessment activities: comparison with standard stereophotogrammetry. Gait & Posture, 39(1), 593NJ. doi:10.1016/j.gaitpost.2013.09.018
[4] Bonnechère, B., & Jansen, B. (2014). Can the KinectTM sensors be used for motion analysis? Transaction on Electrical and Electronic Circuits and Systems, 4(1), 1Lj.
|
|
