Computational and neural mechanisms of divided attention in vision

University Of Washington, Seattle WA

Investigators

Abstract

Project Summary Humans attempt to divide their attention across multiple stimuli many times a day, and unlike our excellent ability to selectively attend to a relevant location in a visual scene, humans are far less successful at doing so. Consider, for example, detecting a traffic light change while changing the radio station in the car. The study of selective attention has made considerable progress over the last few decades but much less is known about the effects of divided attention. Many questions still exist, such as: How many stimuli can you attend to without impairing performance? Under what circumstances can you process multiple stimuli in parallel? Existing behavioral evidence is mixed. It is generally agreed that for simple detection of visual features there is little to no cost of dividing attention. However, as the task becomes more complex, for example detecting a change between two displays, or certain visual stimuli such as words, there appears to be a cost to attending multiple relevant locations. These task and stimuli differences have not been fully characterized and the underlying neuronal mechanisms are unknown. Here, we propose a series of psychophysical and fMRI studies to investigate the role of divided spatial attention on tasks involving visual stimuli. By varying the task demands we can distinguish between a number of candidate processing stages which may be contributing to the cost of dividing attention (Specific Aim 1 and 2). One possibility is that under certain conditions, processing stimulus features is limited at the perceptual level. Another is that holding and comparing multiple visual stimuli in memory introduces errors. A third alternative is that attending to multiple stimuli adds noise at a decision stage of processing. Once we have behaviorally isolated the effects and combinations of different processing stages we will use fMRI to localize neural mechanisms in the early visual areas (Specific Aim 3). Understanding the brain mechanisms responsible for behavioral differences in dividing attention has clinical importance when we consider how many have attention differences as a phenotype. For example, individuals with autism spectrum disorder have difficulties attending to more than one stimuli, while those with ADHD are more likely to have difficulty maintaining attention on multiple stimuli. The underlying mechanisms may prove critical in linking these behaviors with underlying neurophysiological mechanisms.