Figure 1.

Neural tuning to unimanual and bimanual hand movement. (a) Participant T5 performed a delayed-movement task. Cursors on a screen prompted T5 to attempt to make concomitant joystick movements. One of three types of movements were cued on each trial: (1) bimanual: both hands, (2) unimanual left: only left (ipsilateral) hand, or (3) unimanual right: only right (contralateral) hand. (b) Matrix of spike rasters of example electrode no. 97 during bimanual movements. Raster plot (i,j) of the matrix corresponds to electrode 97’s response to right hand movement in direction i while the left hand moved in direction j (colored by right hand direction). Each row of a raster plot represents a trial, and each column is a millisecond time-bin. A dot indicates a threshold crossing spike at the corresponding trial’s time-bin. Different spiking activity can be seen for different bimanual movements, indicating tuning to bimanual movement direction. (c) Tuning curves of example electrodes show a range of tuning changes to each hand (rows) across movement contexts (red/blue). Solid dots indicate the mean firing rates (zero-centered) for movements in the directions indicated on the x-axes. Spikes were binned (20-ms bins) and averaged within a 300–700 ms window after the ‘go’ cue. Shaded areas are 95% CIs (computed via bootstrap resampling). Electrode no. 97 retained tuning for both hands between contexts, electrode no. 13 had suppressed tuning for both hands during bimanual movement, and electrode no. 23 had suppression in left hand tuning during bimanual movement.