Neural mechanisms of background and velocity effects in smooth pursuit eye movements

Abstract

Smooth pursuit eye movements (SPEM) are essential to guide behaviour in complex visual environments. SPEM accuracy is known to be degraded by the presence of a structured visual background and at higher target velocities. The aim of this preregistered study was to investigate the neural mechanisms of these robust behavioural effects. N = 33 participants performed a SPEM task with two background conditions (present and absent) at two target velocities (0.4 and 0.6 Hz). Eye movement and BOLD data were collected simultaneously. Both the presence of a structured background and faster target velocity decreased pursuit gain and increased catch-up saccade rate. Faster targets additionally increased position error. Higher BOLD response with background was found in extensive clusters in visual, parietal, and frontal areas (including the medial frontal eye fields; FEF) partially overlapping with the known SPEM network. Faster targets were associated with higher BOLD response in visual cortex and left lateral FEF. Task-based functional connectivity analyses (psychophysiological interactions; PPI) largely replicated previous results in the basic SPEM network but did not yield additional information regarding the neural underpinnings of the background and velocity effects. The results show that the presentation of visual background stimuli during SPEM induces activity in a widespread visuo-parieto-frontal network including areas contributing to cognitive aspects of oculomotor control such as medial FEF, whereas the response to higher target velocity involves visual and motor areas such as lateral FEF. Therefore, we were able to propose for the first time different functions of the medial and lateral FEF during SPEM.

Keywords: background; distractor; fMRI; frontal eye fields; functional connectivity; smooth pursuit.

FIGURE 1

Schematic depiction of the task. Smooth pursuit eye movement (SPEM) blocks (20 s) alternated with fixation blocks (10 s). The target moved at one of two different velocities in a sinusoidal velocity pattern (0.4 vs. 0.6 Hz) and was presented on a blank screen or with background stimuli. Order of the conditions in the SPEM blocks was randomized

FIGURE 2

Effects of background and target velocity on smooth pursuit gain (in %; panel a), root mean square error (RMSE; in °; panel b), and catch‐up saccade rate (N/s; panel c)

FIGURE 3

Blood oxygen level‐dependent (BOLD) response during smooth pursuit. Results of one‐sample t‐tests (random effects) of the SPEM versus fixation (green), background versus no background (blue) and high versus low target velocity (red) conditions, respectively. Results are reported whole‐brain family‐wise error rate (FWE) corrected voxelwise (p < .05, peak level) with an additional minimum cluster size threshold of 10 voxels. The left hemisphere is depicted on the left. Upper labels refer to the z coordinate of the slices (Montreal Neurological Institute [MNI] space). Significant clusters are binarized.