Abstract
Prior to a saccade, some neurons shift their receptive fields (RFs) to the location that the RF will be in after the saccade (the future field, FF). Current debate focuses on the question whether information about the content of the visual scene in the FF is transferred (feature remapping), or that the transfer is limited to information on the location of salient regions in the visual field (salience remapping). We recorded from neurons in the medial superior temporal (MST) area of one macaque while he was making saccades between two dots. Two patches with independent random motion sequences were present on the screen. One was in the receptive field, the other in a location that would be in the RF after the saccade, but on the periphery, or outside the RF before the saccade (FF). We used a general linear modeling approach to determine the direction tuning in both the receptive and the future field and compare the tuning properties during steady fixation with those obtained pre-saccadically. We uncovered systematic effects of impending saccades on the direction tuning dynamics. Pre-saccadic tuning was often weaker than fixation tuning in the receptive field; this could be related to the behavioral phenomenon of saccadic suppression. Most interestingly we found pre-saccadic direction tuning in locations that were untuned during fixation (i.e. in the FF) as well as a more general pre-saccadic increase in tuning in the periphery of the RF. To our knowledge this is the first electrophysiological evidence that remapped responses are tuned, suggesting that remapping not only transfers salience, but also feature information. This transfer of low-level visual information just prior to a saccade gives the neuron a head start in processing the information in its future field, which could subserve the construction of perceptual stability.
Meeting abstract presented at VSS 2012