Neural decoding of robot-assisted gait during rehabilitation after stroke

Objective·. Recent advances in robot-assisted gait rehabilitation and brain-machine interfaces (BMI) promise to increase the quality of life of people with deficits in bipedal locomotion such as in stroke survivors. Moreover, BMI systems may enhance stroke neurorehabilitation by engaging the user while providing valuable information about cortical adaptations and neural plasticity resulting from robot use. However, the feasibility of decoding walking from brain activity of stroke survivors during robot-assisted gait therapy is unknown. To address this gap, we designed a study to investigate the feasibility of decoding gait kinematics from chronic stroke patients undergoing gait rehabilitation based on a lower extremity gait system (H2 NeuroExo) integrated with a noninvasive neural interface based on electroencephalography (EEG).