EEG decoding of pre-saccadic effects on post-saccadic processing

摘要

Using electroencephalography (EEG) and multivariate pattern analysis (MVPA) we examined whether saccade target information affects post-saccadic target processing. Subjects were instructed to saccade toward a face or a house that, on different trials, could remain the same, change, or disappear during the saccade. We used MVPA to decode between face and house stimuli in the post-saccadic period. The classifier was trained on a separate set of trials without a saccade, where a house or face was presented at the fovea. When the saccade target remained the same during the saccade, we were able to decode the target 154 ms after saccade onset (p 0.05 FDR corrected). In contrast, when the saccade target was changed during the saccade, the new target was decoded at a later time-point, 190 ms after saccade onset (p 0.05 FDR corrected). The 36 ms advantage for the "same" condition suggests that pre-saccadic information facilitated the processing of the post-saccadic stimulus. Finally, the saccade target could be decoded at 235 ms when it had been removed during the saccade (p 0.05 FDR corrected). This result may be a neural correlate of the previously reported "ghost" illusion (Wolf et al. 1980) whereby subjects report briefly seeing the saccade target even though it is no longer present when saccade lands. The late classification may indicate that this illusory percept has a long time course with lower signal strength that reaches recognition threshold later than the physically present stimuli. These findings indicate that information about the (peripheral) pre-saccadic stimulus is transferred across the saccade so that it becomes available and influences processing at a new retinal position (the fovea) when the saccade has landed.