Brain cells become more discriminating when they work together

Team work is just as important in your brain as it is on the playing field: A new study published online on April 19 by the Proceedings of the National Academy of Sciences reports that groups of brain cells can substantially improve their ability to discriminate between different orientations of simple visual patterns by synchronizing their electrical activity.

The paper, “Cooperative synchronized assemblies enhance orientation discrimination,” by Vanderbilt professor of biomedical engineering A. B. Bonds with graduate students Jason Samonds and Heather A. Brown and research associate John D. Allison provides some of the first solid evidence that the exact timing of the tiny electrical spikes produced by neurons plays an important role in brain functioning. Since the discovery of alpha waves in 1929, experts have known that neurons in different parts of the brain periodically coordinate their activity with their neighbors. Despite a variety of theories, however, scientists have not been able to determine whether this “neuronal synchrony” has a functional role or if it is just a by-product of the brain’s electrical activity.

Until recently studies have focused on the firing rate of brain cells as the basic unit of information – the bits and bytes – used by our organic computer. The reason for this fixation was evidence that the firing rates of sensory neurons contain important information. For example, the higher the firing rate of the pain-sensing neurons in the back of your hand, the greater your brain’s perception of pain in that location.

“We are exploring how information is represented by the brain,” says A. B. Bonds, professor of biomedical engineering at Vanderbilt, who co-authored the new study with graduate students “One representation is the firing rate of individual nerve cells, but this does not acknowledge the intricate network structure of the brain, where each cell is connected with 1,000 other cells, on average. One way of representing information that depends on this network structure is the degree of ’agreement’ between groups of brain cells. That is what we have found in the form of the synchronous behavior of groups of cells.”

“For the last five years or so, a growing number of people have been exploring the theoretical possibility that the timing of the arrival of electrical spikes is useful for performing neural computations,” says David Noelle, assistant professor of computer science and psychology at Vanderbilt, who did not participate in the study. “The Bonds paper can be seen as the first step towards providing a test of these theoretical models.”

Scientists studying vision have known for some time that different groups of neurons in the visual cortex respond selectively to the way in which objects are oriented. For example, when a subject views a horizontal bar, one group of neurons begins firing, while a different group begins firing when the bar becomes vertical.

“People have the ability to discriminate between orientations that differ by only a third of a degree. That is pretty remarkable when you consider that individual neurons normally don’t respond to changes in orientation of ten degrees or more! It is even more amazing when you stop to think that a neuron is basically a little sack of salt water!” Bonds exclaims.

Until recently, attempts to study interactions between groups of neurons have been hindered by the fact that researchers were limited to using single microelectrodes to measure electrical activity. Although this technique does a superb job of recording the electrical activity of one or two neurons, attempts to use it to record the activity of a larger number of neurons at the same time has had limited success. (The other method major method for mapping brain activity – functional MRI – measures chemical changes in the brain, not electrical ones, so cannot be used for this purpose.)

Samonds and Bonds used a new technology that employs an array of 100 microelectrodes, which can monitor the activity of dozens of neurons at a time. The researchers used this array to monitor the activity of neurons in the visual center of heavily anaesthetized cats. (The same basic technology was recently approved for clinical trials in paralyzed patients. The goal is to determine if the chip can be implanted in a way that will allow them to control a computer with their thoughts.)

When the subjects were presented with grid patterns at different orientations, the researchers found that groups of about six neurons would synchronize their firing rate for different orientations and that these groups exhibited an ability to discriminate between variations in orientation as small as two degrees, about five times better than individual cells.

“The size of the groups that we can observe was limited by the size of the electrode array,” says Samonds. “Currently, we can monitor about 50 neurons at a time. Synchronization among larger groups should allow higher levels of precision, but we don’t have enough data yet to predict the number of cells necessary to achieve the level of discrimination that many animals possess.”

The researchers also have found that gamma waves – the 30 to 60 Hertz waves that appear everywhere in the brain – may play a key role in this mechanism. They have determined that the neurons respond to a new pattern by synchronizing their activity. When the gamma oscillations are present, the synchronization is maintained 100 percent of the time, but when gamma waves are not present the synchronization breaks down within a few seconds.