Abstract
It has been suggested that first-order motion mechanisms are low-level, reflexive and luminance-driven systems while third order/feature-tracking mechanisms are high-level, salience-driven systems. It is unclear which mechanism is responsible for sensing interocular motion, in which frames are presented to different eyes. In this study we used the Eye Movement Correlogram (EMC) technique to investigate reflexive and attention-driven contributions to pursuit eye movement responses made to motion stimuli that were first-order (binocular luminance sine gratings), third-order (disparity-defined randot square wave gratings), or interocular (spatio-temporal quadrature phase disparity “Shadlen-Carney” sine gratings). Methods: The stimuli were 0.4-cpd horizontal gratings presented with random vertical motion trajectories. Subjects either TRACKED or IGNORED the grating motion while eye movements were recorded with a SRI dual-Purkinje image eye tracker. The eye position was differentiated to derive velocity. The saccade-free eye velocity signal was cross-correlated with stimulus velocity to produce an eye movement correlogram (EMC) for each trial. Results: TRACKING the first-order and interocular gratings produced two-component EMC profiles that comprised an early component peaking between 130 and 150 ms and a late component peaking at 300ms. However, TRACKING of the third-order grating produced a unimodal EMC profile that peaked at 300ms. IGNORING the motion of the first order and interocular gratings produced single-component profiles that peaked early at 120ms. The third-order grating failed to show any significant correlation when IGNORED. Conclusions: Pursuit EMC profiles support the existence of first-order type processing of interocular motion, although about 4 octaves weaker compared to first-order processes.
NIH R01-EY12986 and Erhardt Fellowship to first author