Learning may unify distant brain regions.

In studies of many animal species, the brain shows signs of exerting less effort as individuals learn to perform simple tasks or to recognize relationships between repeatedly presented items. How this apparent neural efficiency arises, not to mention how it facilitates learning, remains unknown.

A new study provides clues to what underlies this efficiency. In laboratory trials, when people first try to discern the locations of various objects, separate brain areas that participate in identifying visible items or specifying their location display characteristic jolts of activity. As people learn an object's location, however, these regions not only toil in a progressively more relaxed style but form a close working relationship that has gone largely unnoted, concludes a team of neuroscientists in the March 5 SCIENCE.

"These two visual pathways in the brain work as one big system in associative learning," says Christian Buchel of the Institute of Neurology in London, lead author of the report. "Reciprocal relationships such as this may promote neural efficiency as organisms learn."

Buchel's team obtained functional magnetic resonance imaging (fMRI) scans of the brains of six adults as they learned the locations of 10 familiar objects displayed one at a time on a computer screen. Each volunteer correctly remembered the location of all 10 items by the end of eight learning trials.

The researchers took a total of 256 fMRI scans per trial for each participant. Scans showed blood-flow rates in the brain, an indirect measure of cells' activity.

The researchers focused on sites located on two previously recognized anatomical pathways in the brain's visual system. Both originate at the back of the brain in tissue that serves as the entry point for visual information. One pathway then runs along the top of the brain and handles object identities; the other takes a lower road and concentrates on object locations.

During the volunteers' early learning trials, blood flow in the two pathways surged. At the same time, a mathematical analysis of fMRI data revealed only a weak relationship between changes in the blood-flow responses along the two routes.

Neural activity slackened in both pathways on later trials, as individuals demonstrated better knowledge of object locations. At the same time, blood-flow changes in the upper and lower pathways became closely aligned.

Buchel and his colleagues view the emergence of this strong link as a sign that the two pathways increasingly pool their efforts during learning trials. Studies of electrical responses in the brain also point to joint efforts among widely separated cell clusters during learning (SN: 2/20/99, p. 122).

The new investigation represents "exciting work," but neural efficiency during learning remains poorly understood, comments neuroscientist Robert Desimoue of the National Institute of Mental Health in Bethesda, Md.

Although most brain-scan studies examine one region at a time, attempts to establish functional relationships between neural regions will rapidly become more common, Desimone predicts.