NIH News Release
National Institute of Mental Health

Thursday, September 5, 2002
2:00 p.m. ET

Jules Asher
NIMH Press Office
(301) 443-4536

Even Neurons Have Favorite Numbers

At least a third of the communicating cells in a front part of the brain critical for reasoning and planning seem adept at keeping track of the number of things seen, report scientists funded by the National Institute of Mental Health. What’s more, they even pick favorite quantities.

Activity of neurons in the prefrontal cortex of monkeys trained to judge the number of dots on a computer screen reflected changes in quantity — peaking for a preferred number and then dropping off progressively as the discrepancy between numbers of dots increased, the researchers found. The study adds to mounting evidence showing how the prefrontal cortex categorizes information from the senses. Drs. Andreas Nieder, David Freedman and Earl Miller, Massachusetts Institute of Technology, report on their findings in the September 6 2002 Science.

“Judging the difference between the number of items we see is one way of categorizing and making sense of them,” explained Miller. “This capacity to quickly glean concepts and meaning from experience underlies the “executive” functions of the prefrontal cortex, which are disturbed in disorders like schizophrenia and autism.”

“The ability to judge the relative quantity of items in the visual field is highly adaptive,” note the researchers. “Social animals such as primates can make decisions to fight or flee by judging the relative number of friends versus foes; in foraging, choosing a larger alternative can contribute to survival.”

Although much more highly developed, the human brain’s capacity to represent numbers likely involves many of the same properties as discovered in the monkey brain, say the researchers. They suggest that future studies may lead to a better understanding of such numerical competence, with possible applications in education and in the treatment of cognitive deficits.

Miller and colleagues are finding that visual categories are represented in the cortex at the level of individual neurons. When monkeys are trained to recognize new categories, neurons in the prefrontal cortex “rewire” themselves to represent those categories.

In the current study, the researchers explored how prefrontal neurons become attuned to quantitative categories. They trained two monkeys to release a lever when the number of dots — from one to five — that appeared on a computer screen matched the number they had seen a second earlier. The animals learned to hold onto the lever if the quantities didn’t match, receiving juice as a reward for correct answers. To make sure the monkeys weren’t cheating — picking up on low-level visual cues instead of actually judging quantity — the researchers randomly varied the locations, sizes and other features of the dots across several trials. While the monkeys were performing this number task, the investigators recorded the electrical activity of 352 randomly selected neurons in the part of the prefrontal cortex where sensory information is processed.

They found that the activity of at least a third of the neurons varied along with the number of dots seen, and their consistent patterns for a given number also generalized across widely varied displays, signaling that the cells were indeed tuning to quantity. Each cell showed a peak of activity for one of the five quantity categories and a systematic trailing-off as the numerical discrepancy between its preferred number of dots and the number presented on a given trial increased.

The researchers speculate that the neurons chose favorite quantities as a result of learning to perform the task. Individual cells become highly specialized as they adapt to experience, participating in multiple networks that can perform many different tasks.

When brain activity for the preferred quantity dropped significantly, there was an increased probability for the monkey to give the wrong answer, providing further confirmation that the measured brain activity was related to their behavior. They made more errors when the difference in the number of dots presented was smaller (e.g., making a comparison between 3 versus 2 dots was more difficult than a comparison between 5 versus 2 dots). Neurons became less precisely tuned — didn’t discriminate as well between quantities of dots — as their preferred quantity increased, paralleling the fact that it’s harder to discriminate between two quantities of equal numerical distance as their magnitude increases. That is, it’s harder to detect a one-dot difference when comparing 5 dots versus 4 dots as opposed to 2 dots versus 1 dot. Overall, the neural activity seemed to form a “bank of overlapping filters” whose properties can explain these observed effects, say the researchers.

NIMH is part of the National Institutes of Health (NIH), the Federal Government's primary agency for biomedical and behavioral research. NIH is a component of the U.S. Department of Health and Human Services.