Abstract
Visual processing is frequently interspersed with saccades, which are associated with strong peri-movement changes in neural and perceptual sensitivity. With fixed gaze, for controlled experiments on processes like covert attention, microsaccades still occur. During the past two decades, it became clear that microsaccades are not random (Hafed and Clark, 2002; Engbert and Kliegl, 2003), but instead exhibit predictable correlations with enhanced or decreased peripheral visual sensitivity (in multiple brain areas) and perception. These time-locked changes appear at eccentricities 1-2 orders of magnitude larger than the eye movement endpoints themselves. However, it remains unclear whether it is microsaccades, perhaps through their motor preparatory activity, that causally influence visual sensitivity, or whether visual sensitivity is itself modulated independently of microsaccades, maybe through oscillatory brain-state fluctuations; in this case, it is such modulations that “leak” into the motor system and trigger microsaccades. Here, motivated by (Hafed and Clark, 2002), in which no attentional effects were present in the absence of microsaccades, we tested the former hypothesis. We used real-time retinal image stabilization to introduce foveal motor errors at fixation (~0.06 deg) and causally drive microsaccades in an experimentally-controlled direction. We then presented peripheral (>5 deg) visual stimuli congruent or incongruent with microsaccade direction and recorded monkey superior colliculus (SC) activity. We found enhanced visual sensitivity and faster reaction times for microsaccade-congruent stimuli as a result of causally-generated microsaccades. In humans, we adapted a known task for pre-saccadic perceptual sensitivity (Rolfs & Carrasco, 2012) for microsaccades: subjects fixated, and we shifted the fixation spot by ~0.18 deg right or left to trigger microsaccades; the actual perceptual task (contrast comparison) was done for stimuli at 7 deg. Peripheral contrast sensitivity was enhanced before experimentally-manipulated microsaccades congruent with stimulus location, consistent with the neurophysiological results. Thus, foveal motor activity is sufficient to influence peripheral visual sensitivity.