Binocular integration of pattern motion signals in the human oculomotor system.

Psychophysical and neurophysiological evidence indicates that the computation of pattern motion is partially, but not completely, disrupted when two drifting gratings forming a plaid are presented dichoptically (i.e., one to each eye, Tailby, Majaj and Movshon, 2010). Here we set out to quantify the extent of this disruption along the visual pathway that underlies short-latency reflexive eye movements in humans. Unikinetic plaids, formed by summing a drifting and a static sinusoidal grating of different orientations, elicit short-latency ocular following responses (OFRs) that can be used to study pattern motion computations (Masson and Castet 2002). In our experiments we combined a vertical drifting (20 Hz) grating and a static oblique grating tilted 45 deg away from vertical (both gratings had a spatial frequency of 0.25 cpd). With this configuration, the magnitude of the vertical component of the OFR is a direct proxy for the strength of the pattern motion signal. We used three stimulus configurations: Binocular - both eyes saw the unikinetic plaid; Monocular - unikinetic plaid in one eye, gray background in the other; Dichoptic - drifting grating in one eye, static grating in the other. We measured, in three subjects, the strength of the vertical component of the OFRs induced by these stimuli. On average, the pattern motion signal for the dichoptic condition was 30% of that generated under binocular presentation, and 52% of that induced by the monocular stimulus. From this latter number we infer that the pattern motion computation is evenly split amongst monocular and binocular pathways, and it is thus likely to rely on computations occurring in primary visual cortex, where both monocular and binocular neurons are present. Meeting abstract presented at VSS 2015.