Abstract
Primates use smooth pursuit and saccadic eye movements to track moving objects. Continuous foveation requires good estimates of target speed and direction. When position and velocity errors get too large, catch-up saccades are triggered. The pursuit system typically reaches an asymptotic direction precision similar to that for perception. We explored the dynamics of directional precision during pursuit initiation and initial saccades. We used a step ramp paradigm to obtain pure pursuit initiation, and a ramp paradigm to elicit an initial saccade followed by pursuit. Small Gaussian blobs moved horizontally across a CRT screen at 10 deg/s in 7 different directions between +/- 10 deg. Instantaneous direction of eye motion at different points in time after pursuit onset were compared to the eye direction obtained for a purely horizontal stimulus. This way, oculometric functions were constructed for direction at each point in time, indicating how well the oculomotor system can discriminate direction. The same calculations were performed for the ramp paradigm, where we also calculated oculomotor direction performance based on the endpoint of the initial saccades. In the step-ramp paradigm, it took about 400 ms after stimulus motion onset for pursuit to reach its minimum threshold of about 2 deg. In the ramp paradigm, initial saccades occurred at 190 ms on average. The saccadic endpoint indicated stimulus direction with a precision of 3.3 deg. At this time, about 100 ms after pursuit onset, pursuit oculometric thresholds for the step-ramp paradigm were still larger than 20 deg and took at least 100 ms longer to reach this value. We conclude that the saccade system reaches more precise estimates of target direction, presumably through the use of positional information that is not available for the pursuit system. Post-saccadic enhancement of eye movement precision might be caused by this saccade-pursuit synergy.
Meeting abstract presented at VSS 2012