Reconstruction of reaching movement trajectories using electrocorticographic signals in humans.

摘要: In this study, we used electrocorticographic (ECoG) signals to extract the onset of arm movement as well velocity hand a function time. ECoG recordings were obtained from three individuals while they performed reaching tasks in left, right and forward directions. The electrodes placed over motor cortex contralateral moving arm. Movement was detected gamma activity with near perfect accuracy (> 98%), multiple linear regression model predict trajectory task three-dimensional space an exceeding 85%. An adaptive selection frequency bands for classification prediction. This demonstrates efficacy developing real-time brain-machine interface movements few eight electrodes.