Figure 1: SSVEP-based passive training on an ankle rehabilitation robot. Subjects sit on a comfortable chair and are requested to gaze at one of the flickering circles. Motion intentions are detected through analyzing SSVEP signals. Each detection would immediately trigger the ankle robot to conduct a predefined passive stretching. (a) An ankle rehabilitation robot is applied to provide passive and active ankle training along DF/PF, INV/EV, and AA trajectories. (b) An EEG system (NT9200, symtop) is applied to acquire and amplify EEG signals.