Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
There is a growing interest for turning biomechanics notably because it is a more challenging task than straight-line walking during which some gait impairments are increased. Detecting heel-strike (HS) and toe-off (TO) events using the trajectory of markers attached to the feet is common in straight-line gait analysis and could reveal very useful to evaluate turning maneuvers. Yet, a comprehensive evaluation is missing, making difficult the selection of features for temporal analysis of turning. This study aimed to compare features of foot marker trajectories to detect HS and TO. Twenty healthy participants, 10 young (5 males, 23 +/- 1 years old, 21.3 +/- 2.2 kg/m(2)) and 10 elderly (4 males, 72 +/- 5 years old, 26.4 +/- 6.4 kg/m(2)), performed quarter, half, and full turns as well as straight-line walking in a gait lab. Fourteen features, adapted from straight-line walking literature, were used to detect HS and TO based on marker trajectories. Force plate measures served as reference. One HS and one TO feature were found particularly suitable. Overall, they detected more than 99% of the 1788 events recorded, with accuracies and precisions of -3.9 ms and 9.0 ms for HS and -7.8 ms and 10.7 ms for TO, respectively. Differences in accuracy and precision were observed among walking conditions and groups, but remained small, generally below 4.0 ms. In conclusion, this study identified kinematic features that can be used to analyze both turning and straight-line walking. Further assessment could be necessary with pathologies inducing severe degradation of gait patterns. (C) 2019 Elsevier Ltd. All rights reserved.
Auke Ijspeert, Grégoire Courtine, Joachim von Zitzewitz, Miroslav Caban, Hari Prasanth, Urs Keller