Submitted for presentation at 2005 meeting of the
Society for Neuroscience
Electrophysiological evidence of intentional suppression in human visual working memory
Y. Yotsumoto1, G.A. Rousselet2; P.J. Bennett2, A.B. Sekuler2, R. Sekuler1
1Volen Center for Complex Systems, Brandeis University, Waltham, MA, USA
2Department of Psychology, McMaster University, Hamilton, ON, Canada
Behavioral studies demonstrate that human subjects can successfully suppress an item in visual working memory. To identify the neural loci and circuitry that support intentional suppression, we measured electroencephalographic (EEG) signals from scalp electrodes while human subjects' recognition memory was tested for either one or two brief, sequentially presented compound sinusoidal gratings that varied in spatial frequency.
As expected, recognition was more accurate when subjects had to remember one rather than two gratings. When subjects were instructed to view but ignore the second if two gratings, suppression was highly successful: recognition was equivlaent to that with just a single item.
Event related potentials (ERPs), were derived from EEG signals time-locked to the onset of the to-be-ignored stimulus. The earliest ERPs to reach occiptal cortex did not differ between ignoring and remembering conditions, showing that suppression does not depend upon down-modulation of early visual processing. However, later ERPs in a left frontal area did differentiate between gnoring and remembering, some 200 msec after stimulus onset; this was followed by analogous ERP differences in parietal cortex at 250 msec, and occipital cortex at 300 msec. Moreover, these late occipital activations were strongly correlated with trial by trial success or failure of memory suppression. The spatial and temporal trajectories of these ERP differences support the idea that inhibitory attentional control by frontal cortex is initiated after the to-be-ignored stimulus has been visually processed, and that processing in visual cortex is later modulated by signals fedback from frontal cortex.
Support Contributed By: NIH, AFOSR, and NSERC