New Approach To Imaging Separates Thought From Perception

"This pre-activation may tune the motion detection pathway, allowing us to respond more strongly and selectively when an anticipated object appears," says Gordon L. Shulman, Ph.D., research scientist in neurology at Washington University School of Medicine in St. Louis.

Shulman and colleagues report their findings in the November issue of the Journal of Neuroscience. They will discuss additional experiments Oct. 24 and 26 during the annual meeting of the Society for Neuroscience in Miami Beach, Fla. "You could use this method anytime you wanted to separate the processes that prepare you to do a task from those involved in executing that task," Shulman says.

A major stumbling block has hampered attempts to understand how cognitive expectations such as goals, memories and thoughts influence the way we see the world. Because images obtained through positron emission tomography (PET) could not be made in less than 40 seconds, they merged expectation with perception. But the advent of functional magnetic resonance imaging (fMRI) cut the time scale to a few seconds. Using BOLD fMRI, Shulman and colleagues have devised a way to obtain separate images of expectation and perception during a single trial. For example, they can obtain one set of images while a person anticipates a visual stimulus and a second set when the stimulus appears.

"This enhanced precision in human neuroimaging allows us to track the sequential involvement of different areas of the brain at different times during a task," says co-author Maurizio Corbetta, M.D., assistant professor of neurology, neurobiology and radiology. "Previously, we could determine which brain areas were active during a task but not the order of their activation."

In the published study, volunteers lay in a scanner and watched white dots on a dark screen. The trials began with a fleeting glimpse of an arrow, which indicated that some of the dots would move in a particular direction (cue period). The subjects then had to press a key if they saw dots moving in that direction among dots that were moving in random directions (test period).

The fMRI method was used to separate areas active during the cue period, in which the subjects interpreted and retained the information about the expected direction of movement, from those active during the test period, in which the subjects used this information to recognize coherent movement of dots.

The region that responded most vigorously during the cue period was a visual area toward the back of the brain called the posterior parietal cortex. This region specializes in spatial perception and eye movements. "These and other data indicate that the posterior parietal cortex plays a major role in funneling our expectations into the visual cortex and influencing perception," Corbetta says. "This conclusion is surprising because experiments with monkeys have attributed this function to the brain's frontal lobe."

Scientists have separated expectation from performance in monkeys by recording from individual neurons. But this approach can monitor only a small part of the brain at any time. At the meeting of the Society for Neuroscience, the researchers will discuss their most recent experiments. They have dissected the cue period further by determining which areas are activated briefly -- these might deduce the direction of the arrow -- and which areas become activated long enough to hold the information on-line until the task begins. "We find that the most sustained time course occurs in the posterior parietal cortex," Shulman says. "So our best guess is that this area holds the instructions."

The researchers also applied their technique to studying the brain's response when a location in space, rather than a direction of motion, was cued. When the subjects were waiting for something to appear in the left visual field, both the left and right posterior parietal regions were activated. But when the object actually appeared at that location, a more ventral parietal region on the right side of the brain became activated. That activation was even stronger when the object appeared on the right but the subjects were expecting it on the left.

This finding may be relevant to stroke patients who have damage to the right ventral parietal region. Such patients often are unable to attend to objects in the left side of the visual field. The Washington University experiments support previous suggestions that the right ventral parietal region may help redirect attention to any place where novel information appears.

"This research sheds light on the causes of attentional deficits after brain injury and could be used to develop new rehabilitation strategies," Corbetta says. "Information about the neural mechanisms of visual attention also may provide insight into the neural mechanisms of awareness, our inner experience of the visual world."

Grants from the National Institutes of Health, the McDonnell Center for Higher Brain Function and the Charles A. Dana Foundation supported this research.

Shulman GL, Ollinger JM, Akbudak E, Conturo TE, Snyder AZ, Petersen SE, Corbetta M. Areas involved in encoding and applying directional expectations to moving objects. Journal of Neuroscience, November 1999.