Preliminary study of Accuracy and reliability of high-speed human-motion tracking using miniature inertial sensors
The purpose of this study was to validate a new motion analysis system which uses the technology of inertial sensors (Moven system). This innovative system is composed of 17 miniature inertial sensors attached to the body. The validation procedure consisted in comparing the data of the right hand trajectory collected by the Moven system with those given by an optoelectronic system (Vicon) composed of 6 infrared camcorders. The comparison of kinematic data was performed during a shot put
... a shot put training session in three high level athletes. The results of this preliminary study only concern the velocity module of the hand (on the middle of the second metacarpal) during the shot put. Paired ttest, r correlation coefficient and Bland & Altman test were used to compare the validity and the limit of agreement between the two tools. The results showed no difference between the measurements (p = 0.066). The correlation between the tools' measurements is r > 0.46. Bland & Altman test showed low reliability (bias < 0.15 m·s -1 ) and accuracy (mean error = 0.63 m·s -1 ) between tools. Similar patterns of the right hand velocity curves during the shot put were obtained with both devices. However, they were differences in velocity intensity for some phases of the throw. The differences in sampling frequency and computational algorithms employed by the Moven and Vicon systems could partly explain these differences in shot-put velocity.