This movie shows real-time 3D control of a virtual robotic arm using epidural electrocorticographic signals. The brain-computer interface uses a co-adaptive algorithm that optimized control by adapting to the subject’s unique cortical signals and performance as they learn to move the virtual arm. Following a short training period, the subject is able to use brain signals to move the arm with balanced speed and precision.

This movie shows chronic (> 4 months) epidural electrocorticographic (ECoG signals recorded from a macaque monkey, The 300,425, 520 and 600 micrometers diameter platinum electrodes were linearly spaced 3mm part (manufactured by J.C Williams’ Lab – University of Wisconsin 0 Madison). Time-domain signals are shown in a 1 s window and at half-speed in order to visualize high-frequency features. The real-time power spectral density of the signal recorded form the 425 micrometer electrode was calculated from a 0.5 s sliding window using a multi taper method.