Abstract
Coherent motion has been used extensively as a research tool to study human motion sensitivity and the properties of motion sensitive neurons, but it has been rarely used to study smooth pursuit eye movements.
Here we measured smooth pursuit eye movements in response to coherent motion. The stimulus consisted of random dots that moved at a speed of 10 deg/sec within a circular aperture of 10 deg radius. The dot density amounted to 2 dots/deg, the dot lifetime was 200 ms. We varied the percentage of coherent moving dots from 20 to 100%. The coherent motion was always horizontal, either leftward or rightward.
In the first experiment, subjects were instructed to initiate smooth pursuit eye movements immediately after motion onset. We found that the amount of coherence had only minor influence on the latency of smooth pursuit initiation, but a strong influence on the acceleration during the initiation period, with faster acceleration for higher coherence levels. Steady-state pursuit gain, however, was similar for all coherence levels. To test if the dependency on coherence is an intrinsic property of smooth pursuit initiation or a result of incomplete motion segregation, we performed a second experiment. Subjects were initially required to hold fixation on a fixation point for 500 ms after motion onset. After fixation offset, they had to pursue the coherent motion for another 1000 ms. Again we found a strong influence of motion coherence: number and magnitude of fixation errors were higher for higher coherence levels. However, we did not find a strong relationship between coherence level and pursuit acceleration after the offset of the fixation point.
The results indicate that smooth pursuit acceleration depends crucially on the segregation of motion signals. If this segregation is completed before pursuit onset, pursuit acceleration is rather immune to noisy motion signals.
This work was supported by the DFG Forschergruppe FOR 560 “Perception and Action”.