As such, most recordings of the human brain are electroencephalograms, more commonly known as EEGs, which use suction-cup electrodes attached to the outside of the head to get a low-resolution measurement of the activity humming inside. But occasionally, the unique opportunity to get a higher-quality electrical recording from a human presents itself. From 2006 to 2008, one such case was available to the laboratory of Nicholas Hatsopoulos, professor of organismal biology and anatomy and chair of the computational neuroscience program at the University of Chicago.
The male subject (whose identity is protected by privacy laws) was part of a clinical trial for BrainGate, a neuroprosthestic device intended to give quadriplegic individuals the ability to move objects with their thoughts. No, this is not a comic book, it actually works – patients in the trial were able to move a computer cursor or, in one case, a robotic arm merely by thinking about hand movement. A grid of 96 electrodes smaller than a Tic-Tac was implanted via brain surgery into the primary motor cortex of the subject, and the electrical signals from those electrodes can then be “decoded” by a computer into movement.
Aside from its prosthetic purpose, the BrainGate implant can also be used as a recording device, one that is many orders of magnitude more specific than an EEG. That allowed Maryam Saleh, a graduate student in Hatsopoulos’ lab, to run an experiment that had never been done before, connecting the activity of specific brain waves in the motor cortex to how closely a person is paying attention to a particular task. The work, co-authored with Jacob Reimer, Richard Penn, Catherine Ojakangas and Hatsopoulos, was published last week in the journal Neuron as a “case study,” a rare scientific study where the n = 1.
“This gave us a really unique opportunity to record, at the micro scale, signals from the human motor cortex,” Hatsopoulos said.
“Brain wave” is a term usually tossed about with little accuracy, but in this case it works: the researchers were measuring electrical oscillations in the motor cortex that roll up and down like a wave. Oscillations at different frequencies (basically the speed with which they go up and down) have been observed in the brain since EEGs were first performed in the early 20th century, and have been associated with various “states of mind” like alertness or deep sleep. More recently, these oscillations have been measured in specific parts of the brain in both monkeys and humans, and associated with attention in sensory systems and the initiation of movement. But attention would also presumably be important in motor cortex, where a quick response to external stimuli is important for a professional athlete, a musician, or someone jumping out of the way of a train.
For this study, the researchers tested their subject using a fairly simple task designed to compare what the brain is doing during attention vs. inattention. The subject was shown a circle surrounded by eight circular targets, and was told to control a cursor with head motions. He was then told that he would be shown five successive instructions of where to move the cursor, but should only follow either the second or fourth instruction (which I’ll call the “true instruction”). The five instructions were shown, then he moved the cursor from the center to the correct target, and the task started over again.
Meanwhile, a slower type of oscillation, called delta, “entrained” itself to the rhythm of the instructions, adjusting its frequency to mirror how often the subject was given an instruction. After the true instruction was given, this entrainment dissipated. That pattern suggested that delta oscillations may function as a sort of “internal metronome,” timing the brain so that it pays maximal attention when it knows it will be receiving important information.
“There are lots of stimuli in the world that have rhythm,” said Reimer, now a post-doctoral researcher at Baylor College of Medicine (and a former science writing co-conspirator with yours truly). “If you’re waiting for a signal that is informative, you could pay attention constantly for a long period of time. But if that thing you’re waiting for has some rhythmicity to it, maybe a more efficient method is to only pay attention ‘on the beat.'”
Taken together, the findings suggest that oscillations at different frequencies play off each other to coordinate the brain’s attention, like the members of a band playing their way through a difficult composition. Under this metaphor, the delta oscillation acts as the drummer, setting the rhythm, while the beta oscillations play off that beat and build to a crescendo when the expected information is presented. That attentional timing could help a tennis player strike a forehand, or a NASCAR driver take the right path on Turn 3 of a racetrack.
Understanding the interplay between oscillations may also help fine-tune the type of brain-computer interface technology that led to the discovery. Rather than sending a constant stream of information from the brain, the computer might be trained to only take action when electrical activity crosses a certain threshold of attention, the authors said.
“The brain-computer interface is meant to help a person move a cursor with his thoughts about movement,” Saleh said. “But when a person is ‘plugged into’ a brain-computer interface, he doesn’t always want to use it; occasionally, he might just want to tune out and do nothing. Using features from these oscillations, the computer can determine when a patient is ready to move.”
More speculatively, Hatsopoulos suggested the oscillations could be used as a way to assess the level of a person’s attention in real time, a tool that any teacher would welcome. Obviously, nobody’s going to be implanting electrode arrays into schoolchildren any time soon, but EEGs may be sufficient. In fact, psychologists have already started to use EEGs to diagnose cases of attention deficit/hyperactivity disorder or post-traumatic stress disorder. As brain-computer interfaces become more common, so too will high-resolution human recordings, allowing scientists to make more and more sense of the brain’s electrical storm.
Saleh, M., Reimer, J., Penn, R., Ojakangas, C., & Hatsopoulos, N. (2010). Fast and Slow Oscillations in Human Primary Motor Cortex Predict Oncoming Behaviorally Relevant Cues Neuron, 65 (4), 461-471 DOI: 10.1016/j.neuron.2010.02.001