Abstract
Background: A previous study of prior probability effects uncovered a mutual interference effect when the two high-probability antisaccade locations were located in the same hemifield, and thus processed by the same hemisphere.
Objective: We hypothesized that a mutual interference effect would have a spatial distribution for targets at different locations within a single hemifield.
Methods: We assessed prosaccades and antisaccades in 8 normal subjects, contrasting low-probability blocks with 8 possible target locations (0.125 probability for each target) and high-probability blocks with 2 possible target locations (0.50 probability for each target). There were four high-probability blocks that varied the angle of separation between the two targets, from 10°, 30°, 90° to 150°. We assessed latency, directional error and amplitude precision.
Results: Effects on prosaccades were minimal. For antisaccades, increasing prior-probability reduced latency and improved accuracy. However, this benefit was less for targets in close proximity (10° or 30°). The patterns of directional errors and amplitude precision for these close-proximity targets showed that saccadic trajectories were deviated away from the location of the other potential target.
Conclusion: The mutual interference effect is strongest when targets are closer than 90° apart, a directional separation that is similar in magnitude to the directional tuning of receptive fields of neurons in the frontal eye field and superior colliculus. Our results also show that this interference is more specifically a mutual repulsion, causing saccades to deviate away from the other location.
supported by CIHR grant MOP-81270