Abstract
Video-based eye tracking is widely used in current vision research, which tracks the pupil center as a proxy to the movement of the eyeball. It has been shown that saccades recorded by such eye trackers may contain post-saccadic oscillations (PSOs), which are positively correlated with the deceleration of the main saccades. However, most research into saccades and PSOs using video-based eye trackers are recorded in cardinal directions only. Oblique direction saccades have been studied mainly by other eye tracking devices, such as scleral search coils. As saccades in cardinal directions are mainly driven in isolation by one of the two pairs of ocular muscles, but oblique saccades are driven by a combination thereof, it stands to reason that different force profiles between cardinal and oblique saccades may lead to different saccadic sequence profiles. We compared the effect of saccade direction on both the main saccadic sequence and PSOs with a video-based eye tracker. Eye movement data was recorded in a dim environment. Participants were asked to make repetitive saccades following a small annulus on a computer screen in eight directions, including horizontal, vertical and obliques. Peak velocity, peak acceleration and peak deceleration of main saccades were compared and PSOs amplitudes were further analyzed and compared with the main sequence parameters. For the main saccades, we find that compared with horizontal and vertical directions, acceleration and deceleration are significantly higher in the oblique directions, although peak velocity is approximately the same. Meanwhile, PSOs amplitude is significantly smaller in the oblique than cardinal directions, even though the deceleration of the main saccadic sequence is higher. Our results show that the movement profiles for oblique saccades may differ from cardinal directions. The relationship for the parameters of the main saccade and the PSOs should be reconsidered in oblique directions.