Abstract
Both attention (stimulus relevance) and expectation (stimulus probability) have been shown to alter information processing in primary visual cortex, suggesting that top-down influences can modulate the very first stage of cortical information processing. However, prior work either confounded attention and expectation, rendering their specific effects unclear, or used fMRI, which has low temporal resolution, leaving it unclear if these effects reflect a modulation of the first feedforward sweep of visual information processing or later, feedback-related activity. The current study orthogonally manipulated stimulus relevance and likelihood while exploiting the high temporal resolution of EEG to investigate if attention and/or expectation can modulate initial afferent activity in V1, as indexed by the early C1 component. Because the C1 is highly variable across individuals, for each participant we first determined two spatial locations at which the C1 could be reliably measured. Next, subjects performed an attentional cuing task in which they were cued on a trial-by-trial basis to direct their attention towards one of these locations and press a button whenever a target stimulus was presented at that location. The probability of a stimulus appearing at a given location was manipulated block-by-block, such that in different blocks the likelihood that a stimulus would appear was high (75%), neutral (50%) or low (25%). ERP analyses revealed that only stimulus probability, not stimulus relevance, may have an effect on the amplitude of the C1, which was larger for predicted stimuli and smaller for unpredicted stimuli, suggesting an effect of expectation, not attention, on the first feed-forward sweep of information processing. Our findings highlight the importance of studies that orthogonally manipulate attention and prediction to determine to what extent modulations of early sensory processing previously attributed to attention, can be ascribed to attention, prediction, or their interaction.
Meeting abstract presented at VSS 2017