Distractor displacements during saccades are reflected in the time-course of saccade curvature

Abstract

Every time we make a saccade we form a prediction about where objects are going to be when the eye lands. This is crucial since the oculomotor system is retinotopically organized and every saccade drastically changes the projection of objects on the retina. We investigated how quickly the oculomotor system accommodates new spatial information when a distractor is displaced during a saccade. Participants performed sequences of horizontal and vertical saccades and oculomotor competition was induced by presenting a task-irrelevant distractor before the first saccade. On half of the trials the distractor remained in the same location after the first saccade and on the other half the distractor moved during the first saccade. Curvature of the second saccade was used to track target-distractor competition. At short intersaccadic intervals, saccades curved away from the original distractor location, confirming that in the oculomotor system spatiotopic representations emerge rapidly and automatically. Approximately 190 ms after the first saccade, second saccades curved away from the new distractor location. These results show that after a saccade the oculomotor system is initially driven by the spatial prediction made before the saccade, but it is able to quickly update these spatial predictions based on new visual information.

Figure 1

The four possible saccade directions and all distractor positions used in Experiments 1 and 2. Participants always made the first saccade to the center dot from either the left or the right side, then the second saccade up or down. Black dots are fixation/saccade targets, white dots are distractors dots, and white arrows illustrate possible saccade directions. The dot sizes and positions are drawn to scale. White arrows are shown for illustration purposes only. Distractor dots were only displaced horizontally and only one distractor was displayed per trial.

Figure 2

An example trial showing the displacement condition in Experiments 1 and 2. In the no displacement condition the distractor stayed at the same location throughout the whole saccade sequence. In Experiment 2 a post-saccadic condition was introduced, which was similar to the displacement condition, except that no distractor was present before the saccade sequence started, but was presented during the first saccade until the end of the trial. The eye represents the current gaze position and the white arrows indicate saccade directions. It should be noted that the figure is not to scale.

Figure 3

(A) Illustration of saccade curvature calculation. The curvature is the median angle α of all angles between each sample point and the straight line between the start and end point of the saccade (red line) which are further away from the start and endpoints than 0.5° of Euclidian distance (black circle around the start and end points). (B) Calculating curvature difference, relative to CW and CCW distractors. The single value curvature difference for the conditions is the difference between the curvature in the CW and CCW trials.

Figure 4

The results of Experiment 1. (A) The position of the distractor before the first saccade and the position of the distractor after the first saccade for each condition. (B) Amount of curvature difference (degrees of arc) as a function of inter-saccadic interval for each condition. Upper panel: Quintile binned data. Lower panel: Data smoothed with a Gaussian kernel, with the black line indicating time points were a weighted within-subjects t-test resulted in p < 0.05 between conditions. In the no displacement condition positive values indicate curvature away from the distractor. In the displacement condition positive values indicate curvature away from the pre-displacement distractor. The shaded areas are 95% within-subjects confidence intervals. Asterisks indicate significant cluster(s).

Figure 5

The results of Experiment 2. (A) The position of the distractor before the first saccade and the position of the distractor after the first saccade for each condition. (B) Amount of curvature difference (degrees of arc) as a function of inter-saccadic interval for each condition. Upper panel: Quintile-binned data. Middle panel: Data from the three conditions smoothed with a Gaussian kernel. Lower panel: Data smoothed with a Gaussian kernel, with the black line indicating time points were a weighted within-subjects t-test resulted in p < 0.05 between the displacement and post-saccadic conditions. In the no displacement condition positive values indicate curvature away from the distractor. In the displacement condition positive values indicate curvature away from the pre-displacement distractor. In the post-saccadic condition negative curvature reflects curvature away from the distractor. The shaded areas are 95% within-subjects confidence intervals. Asterisks indicate significant cluster(s).

Figure 6

Results for the no displacement and displacement conditions after merging the data from Experiment 1 and 2. (A) The position of the distractor before the first saccade and the position of the distractor after the first saccade for each condition. (B) Amount of curvature difference (degrees of arc) as a function of intersaccadic interval for each condition. Upper panel: Data smoothed with a Gaussian kernel, with the black line indicating time points were a weighted within-subjects t-test resulted in p < 0.05 between conditions. Lower panel: Data smoothed with a Gaussian kernel, with the thick blue line indicating time points were a weighted t-test resulted in p < 0.05 from zero for the no displacement condition and with the thick red line indicating time points were a weighted t-test resulted in p < 0.05 from zero for the displacement condition. In the no displacement condition positive values indicate curvature away from the distractor. In the displacement condition positive values indicate curvature away from the pre-displacement distractor. The shaded areas are 95% within-subjects confidence intervals. Asterisks indicate significant cluster(s).