MUFOCA · The behavioral, fMRI, and EEG profiles of multifocal attention

The behavioral, fMRI, and EEG profiles of multifocal attention

Attention allows us to select information from cluttered, visual scenes. By allocating attention to a particular location or feature in the visual field, visual processing of the attended location or feature is enhanced. Behaviorally this leads to faster reaction times and higher accuracy in detecting or discriminating attended objects relative to unattended objects. On the neural level, attention enhances the cortical processing of attended stimuli relative to unattended stimuli. The current project aimed at characterizing and understanding the abilities and limits of visual attention across different types of stimuli. In a first study, the researcher examined the limits of dynamic spatial attention by asking participants to attend to multiple moving objects at the same time. It is well known that participants can typically only keep track of a small number of objects at the same time. For effective stimulus selection, attention often modulates the earliest stages of visual processing, so low-level visual cortex. However, low-level visual cortex is divided at the hemifield, with the left visual cortex processing the right visual half-field, and the right visual cortex processing the left visual half-field. Previous studies showed that the capacity to attend to moving objects is independent for each visual half-field, such that double the number of targets can be tracked when they are presented in separate hemifields (left, right) than when they are presented in the same visual hemifield (e.g., only left; Alvarez & Cavanagh, 2004). In this project, the researcher showed that these behavioral differences can be explained by differences in attentional enhancement of the stimuli themselves. In particular, using EEG, it was found that attention effectively enhances target presentations for objects presented in separate visual fields, but not when objects are presented in the same visual field (Störmer, Alvarez, Cavanagh, 2014). These results indicate that attention – a system usually considered high-level – is constrained by the neural architecture of the perceptual systems, in this case the visual cortex. Following the general hypothesis that limits of attention are determined by the visual representations over which attention operates, in a second study, the researcher examined the mechanisms of attending to features (instead of locations). In particular, participants were asked to attend to different arrays of colored dots (e.g., red) among other-colored dots (e.g., blue) and detect a small motion target. While participants attended to certain colors, visual processing of these colors was measured at other, task-irrelevant locations using EEG. By presenting colors that either matched the target color, or varied in small steps in color similarity from the target color, making it possible to measure visual processing inside and outside the focus of attention in feature (color) space. It was found that when attending to a particular color, for example red, processing of items in that color is enhanced throughout the visual field (global excitation), and at the same time processing of colors perceptually similar to the attended color (e.g., orange) is inhibited (Störmer & Alvarez, 2014). Thus, feature-based attention elicits an excitatory peak at the attended feature with a narrow inhibitory surround in feature space to suppress potentially confusable stimuli during visual perception. This center-surround profile is similar to what other research has reported for spatial attention, and thus suggests that surround-suppression is a canonical principle of attention a selection that can operate based on location or based on features, depending on the task and environment. In a third study, the researcher expanded in characterizing the effects of attention on visual processing by asking whether and to what extend attention modulates processing in higher-level visual regions – namely for complex object categories that exist in the real world. She focused on the processing of faces – a well-learned object category that requires the holistic processing of multiple features. In particular, this study tested the hypothesis that when attending to the category of faces at one particular location, attention to faces spreads globally across the visual field, to other task-irrelevant locations. When attending to simple features such as a single color, attention does not only enhance processing of that color at that particular location, but spreads globally across the visual field (see above). Here the question was whether attention to high-level categories shows similar effects of location independence, or whether attention to a high-level category only affects visual processing locally, within the attended region. The results of this study showed that attention to faces spreads globally – even to regions of the visual field that are currently task-irrelevant and unattended (Störmer, Cohen, Alvarez, in prep). This suggests that attention to simple features such as a color and attention to high-level object categories, like faces, operates similarly by tuning the visual system to the attended feature or high-level object and enhances processing of the attended objects in a spatially global manner. Together, these research questions show the common principles of how attention affects visual processing at different levels of the hierarchy of the visual system. Throughout the project, the researcher studied different types of attention: selection of locations, simple features, and high-level object categories. By examining how attention affects visual processing at these different stages she was able to reveal the common computational principles attention draws on across these domains.

Data: CORDIS, © European Union