Abstract des Autors

The continuous sensing of running biomechanics provides an opportunity to monitor changes in sporting technique for performance or injury prevention. Inertial sensors are now small enough to integrate into footwear, providing a potential platform for continuous monitoring that does not require additional components to be worn by the athlete and that can be used to assess foot kinematics during running as well as temporal parameters. While temporal parameters of gait are already widely used, they may be combined with the measurement of foot kinematics assessed using a wearable, IMU based, foot mounted Inertial Navigation System. Assessment of foot pose at times of foot-ground interaction (such as heel-strike and toe-off) is likely to require excellent accuracy in the face of changing technique and speed. We present and evaluate a threshold free algorithm for assessing temporal gait parameters using a foot mounted IMU. We also investigate the impact of errors in temporal gait parameters on the measurement of foot kinematics at these time points. We find that our algorithm has good accuracy, for example we find a mean error 0.47 +/- 3.84ms for toe-off estimation at a running velocity of 3.4ms−1 . We also find that the magnitude of this error has little affect on some spatial parameter measurements, such as frontal plane foot angle mean error 0.93 +/- 2.07◦ at 2.3ms−1 ). However, for others the error in temporal parameters produces larger changes, for example sagittal plane foot angle (mean error 4.11◦ +/- 3.70 at 2.3ms−1 ).