Toward the use of wearable inertial sensors to train gait in subjects with movement disorders

This study presents an algorithm to estimate gait features, from on-body mounted inertial sensors, and preliminary results on two sampling populations. Estimated gait features are compared with gold standard system (camera-based) for gait analysis. Difference of estimated step length and god standard measure is below 5% when considering median values. Results are promising toward the aim of including this approach in a system for training and assessment of gait for people with movement disorders.