Accuracy of Postural Human-motion Tracking Using Miniature Inertial Sensors. (2016)

• Record Type: Journal Article
• Authors: Dinu, Daniel; Fayolas, Martin; Jacquet, Marine; Leguy, Elsa; Slavinski, Jean; Houel, Nicolas
• Is Part Of: Procedia engineering. Volume 147(2016)
• Page Start: 655
• Page End: 658
• Publication Date: 2016
• Subjects: MVN Biomech -- 3D kinematic analysis -- posture -- accuracy -- centre of mass; Engineering -- Congresses; Engineering -- Periodicals; Engineering; Conference proceedings; Periodicals; 620.005
• DOI: 10.1016/j.proeng.2016.06.266
• Languages: English
• ISSNs: 1877-7058

Abstract

Abstract: Balance control in upright position has direct impact on shooting performance in various competitive activities like archery, rifle shooting, etc. Feet positions and global posture of the athlete change between each shooting activities. Information and short time feedback about the variation of the centre of mass of the athlete during the time before the shoot appears essential for learning and developing skills in shooting sports. In this aim, the purpose of the study was to explore the accuracy and reliability of a motion analysis system which uses the technology of inertial sensors (MVN Biomech system, Xsens). The MVN Biomech system is composed of 17 miniature inertial centrals attached to the full body. The validation procedure consisted in comparing the data of twenty subject's centre of mass directly extracted from the MVN Biomech system with those given by an optoelectronic system (Vicon) composed of 8 infrared camcorders. The centre of mass computed by both system was calculated using the same anthropometric model. For each subject, the comparison of mean position of the centre of mass was performed in three quiet standing of 30 s. Paired t-test, r correlation coefficient test and root mean square (RMS) were used to compare the accuracy between both devices. The results of paired t-test showed a significant difference between the measurements (p < 0.0001) in each component of the mean centre of mass position. Correlation between the tool's measurements was significant and better than r > 0.99 on each component. RMS computation shows mean difference between tools equal to 5.45 mm on X component, 3.25 mm on Y component and 0.73 mm on Z component. The result of the MVN Biomech system appears accurate comparing to the optoelectronic system. The small differences could be explained by the relative motion of the respective sensors and markers of both systems on the soft tissues of the subject. If the MVN Biomech system was preliminary developed to explore the subject's motion, the result of the present study showed that this system could also be used in order to estimate posture or micro-movements like postural sway during shooting activities.