Foot contact detection through pressure insoles for the estimation of external forces and moments: application to running and walking

In motion analysis studies, classical inverse dynamics methods require knowledge of the ground reaction forces and moments (GRF&M) to compute internal forces. Force platforms are considered as the gold standard to measure GRF&M applied to the feet. Such devices reduce the ecological aspect of the experimental conditions by limiting the analysis area. Estimating external forces from motion data and dynamic equations circumvents this limitation at the expense of accuracy. In such an estimation method, the inverse dynamics problem is undetermined since contact is modelled by multiple points representing the potential ground-foot contact area. The contact is systematically multiple during double support phases and using multi-point models that Dorn T. W. et al. (2010) recommends. An optimization approach distributes the forces preserving the global equilibrium on the active contact points according to physiological assumptions, e.g. minimizing external forces. A contact point is considered active when that point on the foot is in contact with the ground. Contact detection is usually based on kinematic parameters such as height and velocity thresholds. Fritz et al. (2019) showed that the tuning of those parameters according to the subject and the task affects the accuracy of the method. Obtaining the correct setting remains time-consuming and requires biomechanical knowledge to be effective. This abstract presents a study that evaluates the potential of pressure insoles to detect contact in an external force estimation method. Two contact detection methods are evaluated: one is based on kinematic thresholds and the other is based on pressure insole data. The evaluation method consists of comparing the GRF&M estimated by both methods with those measured by the force platforms during running and walking.