Visual Attention

Vision is an active process, and our visual experience is dominated by attention, the way we select the most interesting or important parts of the visual scene and dedicate enhanced perceptual resources to these key features. Visual attention can be either overt (when we make an eye movement to an object of interest) or covert (when we attend to something without looking directly at it).

How are different types of attention related, and how do they influence perception?

What are the rules and mechanisms by which attention routes visual information in the brain?

We studied these questions in non-human primates trained to perform a visual selective attention task. We first asked: are the mechanisms for controlling overt attention the same as those that control covert attention? By recording the electrical activity of individual neurons in a part of visual cortex (extrastriate area V4), we found that when we behaviorally dissociated overt from covert attention so that subjects did one without the other and vice versa, both behaviors drove an enhancement of the activity of these neurons (Steinmetz and Moore, Neuron, 2014). From this, we concluded that the mechanisms for controlling overt and covert attention are linked at the level of the enhancement of visual information. In another study, led by Marc Zirnsak, we examined carefully the way that representations of the visual world change in frontal cortex around the time of eye movements, revealing a different pattern of changes than previously believed, a pattern similar to a shift of attention toward the focus of the eye movement (Zirnsak et al., Nature, 2014). These results clarify the nature of the brain circuits controlling overt and covert attention, behaviors that govern our perception of the visual world.

In related work, we studied the way that cortical dyanamics change during attention (Engel*, Steinmetz*, et al., Science, 2016). See here for further description of this work.

Spiking activity of an example neuron, showing similar changes in firing rate when subjects attend to the receptive field stimulus as when they prepare a saccade to the stimulus, on separate trials.