Nicolelis and his colleagues tackled the question by implanting electrodes in a monkey’s brain. These electrodes eavesdropped on dozens of brain cells, called neurons. The researchers then created a computer model that predicted how the monkey would move every time those neurons fired. Once they hooked the computer model up to a robot arm, the researchers could tell how well the model was working. When monkey and robot moved the same way, the model was accurate. It wasn’t an easy task. John Donoghue, chair of the neuroscience department at Brown University, explains that scientists have only recently learned to “listen” to many neurons at once. “We used to listen to one neuron at a time,” he says. “But that’s like trying to understand an entire symphony by listening to the second violinist.”
Donoghue’s lab is one of several working with models that translate brainwaves into movement. He says that scientists still aren’t even sure what they’re hearing when they listen to neurons. While some researchers think that the brain buzzes with useful information buried under background noise, Donoghue thinks that “noise” actually contains cues about how a limb will move. And Andrew Schwartz, a senior fellow at the Neuroscience Institute in San Diego, explains that certain neurons might fire more rapidly when the brain plans movement in one direction or another. While scientists still disagree about these ideas, Schwartz says that new techniques will help solve these and bigger puzzles. “There’s a lot of basic questions we can’t answer, like, ‘How do thinking and memory take place?’ Unless we record from many neurons simultaneously, we won’t be able to unlock these operations.”
Meanwhile, other researchers are already using brainwaves to help paralyzed people communicate. Philip Kennedy of Neural Signals, Inc., has pioneered a technique that lets people with Lou Gehrig’s disease control a computer cursor through electrodes implanted in their brain tissue. Right now, one patient is using the technique, and Kennedy has funding to help six more patients. And Nicolelis predicts that other advances will follow. Several labs are working on chips that may eventually allow patients’ brains to control robots through radio transmitters. Nicolelis also says doctors may one day replace injured nerve cells with silicon chips.
Those advances are still many years away. Within a decade, however, paralyzed patients may control robot arms through computer hookups. That could be the first of many advances that come as scientists piece together the symphony in the brain.
