“School of Cognitive”
Back to Papers HomeBack to Papers of School of Cognitive
| Paper IPM / Cognitive / 14473 |
|
||||||||||
| Abstract: | |||||||||||
|
In order to collect information about the visual world, we move our eyes several times a second; the endpoints of these eye movements are precisely guided by the activity of a network of oculomotor cortical and subcortical areas. Frontal eye field (FEF) is a prefrontal cortex heavily implicated in the control of gaze. The goal of this study is to understand how neural activity in this area drives the accurate execution of saccadic eye movements. Specifically we wish to determine to what degree activity of various functional types of neurons and their interaction within this area contribute to guiding the gaze. We simultaneously recorded FEF activity from multiple neurons with a 16-channel linear electrode during a memory-guided saccade task, and use this activity to predict the saccade endpoints and accuracy. The monkeys were trained to make eye movements toward a previously presented visual target after a 1-second delay (delay period). The target could appear in one of N locations (N=2, 8, or 16) around the fixation point for 1 sec (cue period). We applied various classifications methods to spiking activity and synchrony between simultaneously recorded neurons to predict saccade endpoints. The results were threefold: First, as expected, we found that FEF population activity alone could predict saccade endpoints around the target location to a good extent. Second, we found that biases in eye position during fixation, cue, and delay periods were informative regarding the upcoming saccade endpoints. Interestingly, spiking activity during the fixation period was predictive of the bias in eye position before and during target presentation, and including this bias could dramatically improve endpoint estimation. Thirdly, the power spectrum of spiking activity and coherence between simultaneously recorded neurons reflected saccade accuracy. We found a decrease in alpha-band power and an increase in gamma-band power associated with more accurate saccades. Noise correlations between FEF neurons during the fixation period also varied with subsequent saccade accuracy, with greater noise correlations preceding less accurate saccades. Our results identify potential neural substrates within FEF underlying the precise control of saccadic eye movements.
Download TeX format |
|||||||||||
| back to top | |||||||||||
