For centuries, magicians and illusionists have known that if they distract your gaze and attention, you are unlikely to notice what they are doing in plain sight. This inattentional blindness is a result of our limited ability to take in everything about the visual scene and is traditionally illustrated in the lab by set-size effects in visual search: the more items you have to search through, the harder it is to find what you are looking for and the longer it takes. We have recently found that this limitation does not appear to involve visual processing in visual areas, but takes place as the visual information is passed up to association cortices. We hypothesized that divisive normalization in the lateral intraparietal area (LIP) of posterior parietal cortex plays a key role in this process. Our hypothesis combines our knowledge of the neurophysiology of attention, divisive normalization and decision making to explain the behavioral effects associated with changes in set-size. Specifically, we have suggested that when looking for an item, activity in LIP represents the accumulation of evidence from visual areas that the item is in the neuron's response field and if the activity reaches a threshold before a deadline, the animal indicates that the item is in that location. This accumulation is affected by divisive normalization: the more items in the visual world, the more activity across the area and the greater the normalization, resulting in lower accumulation rates. We know that visual attention enhances responses in visual areas, and propose that this effective increase in gain in the input to LIP helps override the effective gain decrease due to the divisive normalization. In this project, we will test this hypothesis using a visual search task utilizing moving dot stimuli. Specifically, we will record from single neurons in areas MT (a motion processing area) and LIP while animals perform a visual search task in which the 1, 2 or 4 objects in the array are moving dot patterns. If a target direction is present, then the animals must look at it, otherwise, they must maintain fixation. We will vary the signal and the color of the dots in each patch and will have three attentional conditions. In the spatial attention condition, the animal is spatially cued, indicating which location the target will appear in, if it appears at all. In the feature- based attention condition, the animal is given a color cue and the target, if it appears, will appear in a patch with dots the same color as the cue. In the spread attention condition, the animal is not given any cue. We will model the behavior (both percent correct and reaction times) based on our hypothesis and the activity in MT and LIP. We will then directly test this hypothesis, by pharmacologically manipulating responses in the frontal eye field while recording behavioral and neuronal data from LIP and MT. We expect that our results will shed light on the neuronal mechanisms underlying our limited ability to process visual information in the scene, but could also indicate a functional role for the attentional modulation that has been studied for almost 30 years.
This project is aimed at studying the neural mechanisms that limit how many things in visual space we can attend to. By understanding these mechanisms, it may be possible to design better visual environments (head up displays, security displays, dashboards, etc) that can help attention stay where it is needed most, particularly in environments where distractions can have life-altering consequences (such as for the TSA, in the military, or even when driving a car on the freeway).
