Abstract
Introduction: Microsaccades are miniature saccades that occur frequently and involuntarily during fixation. Microsaccades cause sudden and rapid changes to the retinal image, but we do not perceive motion or blur. Both psychophysical performance and neural activity are transiently suppressed by large, voluntary saccades, a phenomenon called saccadic suppression. This study tested whether cortical activity, measured with fMRI in humans, is likewise suppressed with microsaccades.
Methods: Saccadic suppression was tested by comparing visually-evoked response amplitudes with and without saccades. In Expt 1, subjects made microsaccades while fixating a stable target. In Expt 2, subjects made large (14 deg) voluntary saccades to a peripheral target. In both experiments, a task-irrelevant stimulus appeared at various intervals relative to saccade onset. fMRI response amplitudes were measured coincident with: 1) saccades that occurred in the absence of a visual stimulus; 2) visual stimuli that occurred in the absence of a saccade; and 3) stimuli with onset times within 100 ms of a saccade. Analysis focused on subregions of V1 that responded to the stimulus in the absence of saccades, but that showed no evidence of saccade-related activity in the absence of visual stimulation.
Results: In Expt 1, responses to the visual stimulus were larger when the stimulus appeared within 100 ms of a microsaccade relative to when the stimulus appeared during stable fixation. This enhancement could not be attributed to the summation of microsaccade-alone and stimulus-alone activity. In Expt 2, visual responses were smaller around the time of a large, voluntary saccade.
Conclusions: Saccadic suppression predicts a decrease in response amplitude, yet we observed enhanced visual activity. This enhancement is not likely due to metabolically-demanding inhibitory activity, because we observed a reduction of visually-driven activity in conjunction with large voluntary saccades. Our results suggest that perisaccadic changes in visual responses depend on eye movement metrics.
Meeting abstract presented at VSS 2012