Skip to main content
Fig. 1 | Journal of NeuroEngineering and Rehabilitation

Fig. 1

From: Automatic versus manual tuning of robot-assisted gait training in people with neurological disorders

Fig. 1

Overview of assistance tuning. The assistance was either AT based on the error between reference and measured trajectories or MT by a therapist. In this figure only an example for the foot clearance subtask is shown, however, the algorithm was applied to all subtasks shown in Table 2 simultaneously. For the AT algorithm, based on the error, every three steps, the assistance was either increased (if error >upper bound, see Table 2), decreased (if error <lower bound) or kept constant (other cases) by scaling the amplitude of the assistance profile (K) shown on the right. For the MT approach, the therapist could change the assistance (amplitude of the assistance profile K on the right) for each subtask by using graphical sliders. Feedback for the therapist was also shown to assist the therapist in tuning the assistance. As shown in this figure, the therapist got feedback about the maximal knee angle for the foot clearance subtask. The purple bars represented the maximal knee flexion angles for the previous three steps of the less impaired leg, while the blue bars represented the maximum knee flexion angles for the more impaired leg. The green line indicated the maximal knee flexion angle for the reference trajectory

Back to article page