A brain-computer interface study reveals one brain region's surprising role in planning movements exclusively in response to sounds.
PROVIDENCE, R.I. [Brown University] - In the swimming pool game Marco Polo, "Marco" navigates toward other players with eyes closed, responding only to hearing the other players say "Polo." Success depends on the ability to move one's body in response to sound cues alone, and a new study finds that a specific part of the brain may help make that possible.
The study, published in the journal Scientific Reports, provides evidence that neurons in the middle frontal gyrus - a part of the brain's frontal lobe - may play a role in planning body movements, but only when those movements are in response to auditory stimuli. The findings represent what could be a previously unknown function for this part of the brain and could provide a new target for researchers developing assistive devices for both movement and hearing disorders.
The work was part of the BrainGate clinical trial, which studies a tiny investigational implant capable of recording information directly from the brain and using that information to drive the movement of computer cursors or even robotic prosthetic devices.
"One of the opportunities afforded by the BrainGate clinical trial is that at the same time as we're working toward helping people with paralysis, we're also learning new things about the human brain," said Dr. Leigh Hochberg, a neurologist, professor of engineering at Brown University and director of the trial and BrainGate consortium. "This finding turned out to be a complete surprise, which is exciting."
Up to now, brain-computer interface (BCI) implants like BrainGate have mostly been placed in the motor cortex, a part of the brain associated with voluntary movement. But researchers are interested in harnessing additional signals for BCIs by exploring other parts of the brain, including areas that may be involved in the upstream planning of movements and actions. It was in the process of looking at a new brain region for BCI recording that the researchers made this new discovery.
"There's a large body of literature to suggest that parts of the premotor cortex toward the front of the brain - the region we were looking at in this study - become active earlier than the more posterior regions of motor cortex during movement tasks," said Carlos Vargas-Irwin, an assistant professor of neuroscience (research) at Brown and study co-senior author. "If we're able to record signals from these areas, it's possible that we could even further speed the responses of our neural interface system. That's what we were investigating."
For the study, a clinical trial participant with paralysis in his arms and legs from a spinal cord injury was asked to perform a simple movement-related task: after observing a shape in the corner of a computer screen, he would attempt to grab it and move it to the middle when cued. This was first done while the participant was being monitored by fMRI, a non-invasive method of studying brain signals in real time. The fMRI data suggested that a specific part of the premotor cortex - located in the middle frontal gyrus - seemed to be active during the task.
The next step was to implant a BrainGate microelectrode array near that part of brain in the same participant and repeat the attempted movement task. But to the researchers' surprise, the array detected no informative signals during a repeat of the movement task.
"We expected, based on the fMRI data, that we'd see some related activity," said Hochberg, who also directs the Veterans Affairs Rehabilitation Research and Development Center for Neurorestoration and Neurotechnology. "So it was puzzling, to say the least, when we didn't."
That's when the study took yet another unexpected turn. When the team looked at data from a related research session - one in which they verbally told the participant which target to reach for - suddenly the array picked up a strong signal from the middle frontal gyrus.
"That's when our puzzlement turned to pleasant surprise," Vargas-Irwin said.
It led the team to design new research that alternated between giving auditory cues and visual-only cues. That work helped to confirm that - at least in this participant - neurons in the middle frontal gyrus produced movement-related signals only in response to auditory clues.
"It was as if these neurons simply weren't interested in visual information," Hochberg said. "But they reliably responded to auditory information, and that was completely unexpected."
The researchers caution that this was research involving a single participant, so more work needs to be done to confirm and generalize the findings. But the study does suggest a new target to be explored in pursuit of improving neural interface devices.
"The decoders we use to convert neural activity into action are focused on activity that's happening in the precentral gyrus right now," Hochberg said. "To complement this, we'd like to train decoders to interpret and harness the activity that happens 200 milliseconds or more before the activity begins in the motor cortex. If we can do that, it could eventually lead to even more reliable and intuitive neural interfaces."
Hochberg, who is affiliated with Brown's Carney Institute for Brain Science, says he's pleased that the BrainGate clinical trial has potentially revealed a new basic neuroscience finding.
"This work involved postdocs and graduate students from neuroscience as well as engineering, and I think it underscores what can be accomplished in a multi-disciplinary research environment," he said. "We're also thankful for all of our remarkable clinical trial participants, who really make this work possible. They join us in this research to help other people with severe speech or motor impairments, and it's only with our participants' incredible dedication, insights and feedback that we can create the neurotechnologies that will restore communication and mobility."
Additional authors from Brown include: Tommy Hosman, Jacqueline Hynes, Jad Sab, Kaitlin Wilcoxen and John Simeral. Authors from Massachusetts General Hospital include Bradley Buchbinder, Nicholas Schmansky, Sydney Cash, Brian Franco and Jessica Kelemen.
The research was supported by U.S. Department of Veterans Affairs (N9228C, N2864C, B6453R, P1155R, A2295R), the National Institutes of Health (R01DC009899, U01DC017844, UH2NS095548, U01NS098968, T32MH020068, DP2 NS111817) and Massachusetts General Hospital.
CAUTION: Investigational Device. Limited by Federal Law to Investigational Use.
