Skip to Main Content
Through certain mental actions, our electroencephalogram (EEG) can be regulated to operate a brain-computer interface (BCI), which translates the EEG patterns into commands that can be used to operate devices such as prostheses. This allows paralyzed persons to gain direct brain control of the paretic limb, which could open up many possibilities for rehabilitative and assistive applications. When using a BCI neuroprosthesis in stroke, one question that has surfaced is whether stroke patients are able to produce a sufficient change in EEG that can be used as a control signal to operate a prosthesis. The aim of this paper is to determine if post-acute (<;;3 months) stroke patients are able to use an electroencephalogram (EEG)-based BCI to trigger neuromuscular electrical stimulation (NMES)-assisted extension of the wrist and fingers. EEG was recorded while subjects performed motor imagery of their paretic limb, and then analyzed to determine the optimal frequency range within the mu-rhythm that showed the greatest attenuation. With the help of visual feedback, subjects then trained to regulate their mu-rhythm EEG to operate the BCI to trigger NMES on their wrist extensor muscles. 9 post-acute (<;;3 months) stroke patients, aged 58.2 ± 9.3 yrs, participated in this study. 4 out of 6 subjects who completed the trial are able to use the BCI to trigger NMES on their paretic wrist extensor muscles. This study presents findings that movement intention, as characterized by the attenuation of mu-rhythm EEG, is detectable in post-acute stroke patients, and that this signal is can be used as a control signal for the patients to operate a BCI to trigger NMES.