The Role Of Visual Feedback For Online Trajectory Corrections And Generation Of Secondary Saccades In The Antisaccade Task

Abstract

Purpose: : Antisaccades are volitional eye movements requiring suppression of a reflexive saccade toward the stimulus (prosaccade) and generation of a saccade toward a location that is the mirror-opposite to where the target was presented (antisaccade). Visual feedback (online control) can be used during an ongoing saccade to minimize the final error. We investigated (1) whether visual feedback presented at the desired antisaccade location enhances online antisaccade trajectory control; and (2) the role of secondary saccades (SS) as a complementary offline strategy to correct primary saccade (PS) amplitude error.

Methods: : Twelve visually normal observers viewed targets presented at ±10° or ±17° horizontally. Eye movements were recorded by video-based eye tracker. Trials were conducted in 2 blocks: prosaccade and antisaccade. Target feedback was provided randomly on 50% of trials in both blocks. The extent of online control was examined by calculating the coefficient of determination (R²), relating the position of the eyes to the endpoint position throughout the trajectory. Low R² values in mid-trajectory indicate error reduction by online control. Frequency and kinematics of SS were examined to investigate their compensatory role.

Results: : Z-scores (transformed R² values) were higher for antisaccades than prosaccades, indicating less online control during antisaccades (p=0.002). Visual feedback did not affect the control of saccade trajectory in either task. However, SS frequency increased with target feedback in both tasks (prosaccade with feedback=48.2% vs no feedback=9.0%; antisaccade with feedback=67.2% vs no feedback=43.2%; p<0.0001). The compensation for amplitude error at the end of the PS (as measured by the mean transformed Pearson r coefficients) was also greater with target feedback in both tasks (prosaccades with feedback=0.96±0.16 vs no feedback=0.62±0.15; antisaccades with feedback=0.74±0.14 vs no feedback=0.21±0.08; p=0.002).

Conclusions: : Visual feedback presented at the antisaccade target goal does not affect online control of saccade trajectory. However, visual feedback plays an important role in generating SS in the antisaccade task. Specifically, the frequency and the ability of SS to correct post-primary saccade error increased when visual feedback was available.

Keywords: eye movements: saccades and pursuits • vision and action • motion-2D