Attention acts as visual glue

When you gaze at a bowl of fruit, why don’t some of the bananas look red, some of the apples look purple and some of the grapes look yellow?

This question isn’t as nonsensical as it may sound. When your brain processes the information coming from your eyes, it stores the information about an object’s shape in one place and information about its color in another. So it’s something of a miracle that the shapes and colors of each fruit are combined seamlessly into distinct objects when you look at them.

Exactly how the brain recombines these different types of visual information after it has broken them apart is called the “binding problem” and is currently the subject of considerable controversy in the neuroscience community. But the results of a brain mapping experiment, published online by the Proceedings of the National Academy of Sciences on July 29, provide significant new support for the theory that attention is the glue that cements visual information together as people scan complex visual scenes.

The study was a collaboration among René Marois, assistant professor of psychology at Vanderbilt; John C. Gore, who recently moved from Yale to become a Chancellor’s University Professor at Vanderbilt; and Yale graduate student Keith M. Shafritz.

“There are more than a dozen places in the brain involved with processing visual information, each specializing in information with slightly different attributes,” says Marois. “Some specialize in processing color, some specialize in processing shape, while others specialize in movement. These areas are not clustered together, but distributed widely around the back of the brain.”

There are two leading theories about how the brain reintegrates this information.

One view proposes that the neurons in the scattered areas are bound together in a way that allows them to act simultaneously. When you look at a banana, the neurons that store information about the banana’s shape fire simultaneously with the neurons in a different region of the brain that store information about the banana’s color. It is the direct functional interaction between neurons located in different visual areas that binds together an object’s numerous visual properties.

In the 1980’s, Anne M. Triesman at Princeton and her colleagues advanced an alternative mechanism. She proposed that visual binding is mediated by the parietal cortex, an area of the brain known to be involved in spatial attention. She suggested that the act of focusing one’s attention on an object’s spatial location provides the key that binds the different types of visual information together. If an apple is sitting on the table in front of a woman, then her brain, specifically the parietal cortex, associates the information about its color and shape with its location and uses the spatial information to bind together the visual information whenever she focuses her attention on the apple.

The description of a patient who, following a brain injury in the parietal lobe, had difficulty associating colors with more than one object at a time gave Marois the idea for the basic experiment. When the person was presented with objects one at a time, he had no problem properly pairing their shapes and colors. When presented with two or more objects at the same time, however, he often mismatched the color of one object with the shape of another.

So Marois designed a series of trials that asked subjects to concentrate on the shape only, the color only or both shape and color of pairs of objects displayed on a computer screen while their brain activity was monitored using the technique called functional MRI. The researchers presented these pairs to the individuals either sequentially in the same location or simultaneously at different locations and recorded the areas in the brain that were most active.

“The purpose of our study was really to test the attention theory as strongly as we could,” says Marois. “I was actually surprised that it worked because we had to adopt such stringent testing conditions.”

Despite their stringency, the tests showed that activity in the parietal region increased significantly whenever the individuals were presented with more than one object at the same time.

“This provides strong evidence in favor of the theory that spatial attention is the binding glue that the brain uses to integrate visual objects whenever it is presented with more than one object at the same time, which is most of the time,” says Marois.

While the study results support the attention theory, they do not rule out other mechanisms. “In fact,” he adds, “it is practically certain that the brain uses several mechanisms to solve this fascinating problem.”

The project was funded by a grant from the National Institute of Neurological Disorders and Stroke.

For more news about research at Vanderbilt, visit Exploration, Vanderbilt’s online research magazine at http://exploration.vanderbilt.edu.