Assessment of EEG event-related desynchronization in stroke survivors performing shoulder-elbow movements

Abstract

It is unknown whether electroencephalography (EEG) signal characteristics in stroke survivors with motor deficits register enough activity for use with brain-computer interfaces (BCIs). This research studied pre-movement EEG from shoulder-elbow movement in stroke survivors to identify signal characteristics potentially useful for robot-assisted stroke rehabilitation. Pre-movement event-related desynchronization (ERD) was examined in the alpha band mu rhythm for control (n = 7) and stroke subjects (n = 11). Subjects were all right-hand dominant; stroke subjects used their impaired arm and controls were assigned a side to match stroke subjects. Both non-dominant-arm-tested stroke and control subjects exhibited greater ERD intensity vs. those using their dominant arm (p < 0.05). Also, pre-movement ERD was detected in stroke survivors, which suggests at the possibility of using ERD as a BCI system control signal. However, the peak ERD of stroke survivors was significantly lower than that of healthy subjects (p < 0.05), which brings doubt to whether the intensity of ERD in stroke survivors is large enough to be used as a BCI system control signal