A pair of Princeton scientists have found that it has a lot to do with the brain circuits that control eye movements. Neuroscientists Tirin Moore and Katherine Armstrong showed that these brain circuits serve a double function: In addition to programming eye movements, they also trigger amplification or suppression of signals that pour in from the locations where the eyes could move.
The finding, published in the Jan. 23 issue of Nature, is the first to pinpoint a neural mechanism behind one of the most fundamental aspects of mental activity -- the ability to direct attention to one thing as opposed to another.
"Without regulating your attention, you would orient to everything that appears and moves. An organism that couldn't filter anything just wouldn't work. It would be in a state of constant distraction," said Moore. "This work shows that, whether we are moving our eyes or not, the networks that control eye movements may be a source of that filtering."
Working with monkeys, the researchers picked a site in the brain area that controls eye movements and established exactly where neurons at that site made the eyes move. They then located a single neuron, in another part of the brain, that was responsible for processing visual stimuli from precisely the same location targeted by neurons at the eye movement site.
With the monkeys trained to fixate on the center of their visual field, the researchers displayed an image in the location associated with the two brain areas. They then electrically stimulated the eye movement neurons, but not strongly enough to actually make the eyes move. When this microstimulation was applied, the visual processing neuron showed a much greater response to the displayed image than when the electrical stimulation was not applied. On the other hand, when no image was being displayed, microstimulation of eye movement neurons had no effect on the visual neuron.
The researchers concluded that the very act of preparing an eye movement to a particular location caused an amplification of signals from that area. These eye movement neurons acted like a volume control on an amplifier, controlling the strength of the signal from one particular spot in space, but not altering the quality of that signal. By stimulating neurons in the eye movement area, the researchers in effect forced the animal to shift its attention from one location to another even though it did not move its eyes.
The study hinges on a long-known fact in visual attention -- that humans and primates can attend to something without moving their eyes to that object. This ability is useful for many animals that encounter social situations in which there is a potential danger in looking directly at another animal. But scientists were unsure how closely eye movements were tied to the phenomenon of attention.
Moore and Armstrong's finding builds on an earlier study in which Moore observed behavioral effects of electrically stimulating eye movement neurons. In that study, monkeys were better able to detect subtle changes in a visual target when their eye movement neurons had been stimulated. The new study, which measured electrical output of visual neurons rather than measuring a behavioral effect, draws a much more powerful conclusion about how the brain is wired.
Calling the study a "landmark," neuroscientist William Newsome of Stanford University compared the work to discovering how the ignition system of a car is wired. "You know, from looking at the car behaviorally, that if you put the key in the ignition and turn the crank it leads to the car starting," said Newsome. "But if you really want to understand what's going on inside that car -- if you want to go in there and fix things when they go wrong -- you need to know how that behavior comes to pass. Where does the signal go? And then where does it go from there?"
There are many human diseases and disorders that involve defects in information processing and attention -- most famously attention deficit disorder -- for which scientists would like a firm idea of what neural circuits are involved, said Newsome.
"It takes the whole attention field and steps it up a notch, because now people can start asking questions about mechanisms," said Michael Shadlen, an expert in visual perception at the University of Washington.
Apart from the particular finding about spatial attention, the study reveals an important technique that could be used to trace many other types of neural circuits, the researchers said. "Short-term memory, decision-making, planning motor acts all involve flow of information from one area to another and until now we have had no real way to monitor that information flow or reproduce it in the laboratory," said Newsome.
A next step, said Moore, will be to further analyze the eye movement neurons and find out whether they act alone in regulating spatial attention. Another experiment would be to see whether manipulating these neurons and ostensibly making an animal attend to one place or another can determine what information the animal remembers. "If you don't attend to something, you don't see it," Moore said. "There are many things that hit our retinas, but we don't experience them and don't remember them unless we pay attention to them."