Abstract
Accurate spatial localization of a visual stimulus requires the integration of its retinal position with the direction of gaze at the time of stimulus presentation. But studies have shown that brief visual targets presented near the time of saccades tend to be mislocalized. Mislocalization starts growing long before a saccade (∼100ms) and reaches a maximum at saccade onset. This presaccadic mislocalization has often been attributed to the temporal mismatch between the visual and eye position signals due to the long visual afferent delay (>50ms). To directly test if the visual afferent delay is necessary for presaccadic mislocalization, we investigated how observers localize phosphenes induced by transcranial magnetic stimulation (TMS) of the visual cortex (thus bypassing the afferent delay) around the time of saccades.
Observers who consistently perceived single phosphene with a sharp vertical edge were screened and tested in two conditions. In one condition, a dual-pulse TMS was applied over the left visual cortex at various times while they were performing a delayed (rightward) saccade task. Observers reported the perceived position of the leftmost edge of the phosphene with a mouse cursor presented after a saccade. In the other condition, instead of applying TMS, a dim vertical bar was presented (13ms) on the monitor screen.
Most observers mislocalized TMS-induced phosphene near the time of saccades as they did for the flashed bar. The time course and magnitude of the error for both types of stimuli were practically identical. Thus presaccadic mislocalization should be explained with some other factors than afferent delay.