Abstract
We investigated how selective attention to a target embedded in a textured background interacts with contrast gain control. Observers attended to an eccentric target surrounded by a concentric annulus that was either coextensive with the target (unsegmented) or segmented from it by the introduction of a small gap, or phase offset. A brief contrast increment appeared on the target on 50% of the trials. Because contrast gain control is phase insensitive, it predicts the same modulation of the surround whether it is in-phase or out-of-phase with the target. However, selective attention predicts that the surround will be more suppressed in the unsegmented condition, resulting in a smaller response at the surround frequency. We used steady state visual evoked potentials (SSVEP) with target and background tagged with different frequencies, to unambiguously monitor the response from target and background. For each observer we collected high-density EEG data from increment-absent trials (n = 7). We then used a minimum norm inverse procedure combined with realistic MR-derived head models and retinotopically-mapped visual areas to estimate cortical activities for target and background. Our preliminary data suggests that the response in primary visual cortex to the surround that is in phase with the target is more suppressed than the surround that is out-of-phase. This finding argues against a purely contrast normalization account of the data. At the same time, the responses evoked by the target remained unchanged across the different conditions. These results suggest that selective attention may involve an active suppression of textured backgrounds that are co-extensive with the target.