Abstract
Perceptual suppression occurs robustly around saccades, but its underlying mechanisms are debated. On the one hand, purely visual masking effects may be sufficient; on the other, pathways for extra-retinal saccade-related corollary discharge exist. However, possible interactions between these two mechanisms remain unexplored. Here we show, using human psychophysics and retinal-circuit electrophysiology, that purely visual suppression[Office1] originates at the very first stage of the visual system in the retina itself, that it has a much longer time course than perceptual effects with real saccades, and that it dictates perceptual dependencies of suppression on image statistics. In the human experiments, 4 subjects located a low-contrast stimulus flashed at different times around saccades, while viewing a patterned background with 1 of 3 possible dominant spatial frequencies. In separate experiments, subjects maintained fixation and the background moved rapidly for 70 ms to "simulate" saccade-associated retinal image shifts. In ex-vivo retinal electrophysiology, we recorded retinal ganglion cell (RGC) activity in isolated mouse and pig retinae using multi-electrode arrays during a comparable simulated saccade paradigm. Critically, the same background and flash manipulations were employed. Perceptually, contrast sensitivity was reduced after both real and simulated saccades, but this reduction lasted significantly longer for simulated saccades. Suppression was also weakest and shortest for high spatial frequency backgrounds regardless of condition. RGC responses to flashes were also strongly modulated after rapid image shifts and with similar dependencies on background image statistics. Critically, RGC suppression lasted for much longer than in both perceptual experiments. We conclude that both visual and corollary discharge mechanisms may interact synergistically during saccadic suppression: retinal-circuit visual effects dictate the overall properties of saccadic suppression, including dependencies on image statistics, and corollary discharge signals instead act to dramatically shorten retinal-circuit masking effects which would otherwise last for up to ~1 second after every saccade.
Meeting abstract presented at VSS 2018