Roadmapping the Adoption of Brain-Machine Interfaces


by Kayt Sukel

April 15, 2013

Imagine a future where a tetraplegic person, one who has lost control of all four limbs, can walk on a pair of robotic legs instead of spending her days in a wheelchair. Or perhaps one where people with Locked-In Syndrome can easily and freely communicate with friends, family, and caregivers. Or even one where people who have had strokes can help their doctors create tailored rehabilitation plans based on how their wounded brains respond to various stimuli.

These may sound something you might find in a popular science fiction novel, but this future may be closer than we think. Neuroscientists, engineers, and computer scientists across the globe are hard at work on clinical brain-machine interfaces (BMI), a new breed of assistive technologies that directly harness the brain’s signals in order to help those with disabilities walk, talk, and participate in other common activities of daily living.

Creating a BMI roadmap

In late February, academics, clinicians, government representatives, and others met in Houston to create a strategic plan for developing and implementing clinical BMI systems. BMI have received a lot of attention in recent months, and in the glare of that spotlight, the community needs to gain consensus on how to safely and effectively bring these new technologies from the bench to the bedside, says Jose L. Contreras-Vidal, chair of the International Workshop on Clinical Brain-Machine Interface Systems.

“There have been some amazing demonstrations of what these kinds of technologies can do, but it’s been unclear how to evaluate these technologies for safety and use,” he says. “This is the time to form a plan, or roadmap, so we can assess these systems properly and, ultimately, bring these technologies out of the lab and into the clinic.”

Speakers at the workshop discussed the history of implantable devices, the ethical considerations of use, what patients need, and what we’ve learned so far from current clinical trials. Even in such a gathering of experts, Contreras-Vidal said that it quickly became clear that the future of clinical BMI systems needs to be patient-centered.

“At the end of the day, it does not matter how good the technology is. You can have the best technology but if the patient doesn’t want to use it or wear it, it all ends there,” he says. “We understand that, we’ll have to have focus groups across the community so we can come up with a set of methods to assess the users and the types of technologies they really need.”

Invasive vs. non-invasive techniques

BrainGate is an ongoing, multi-center clinical trial evaluating the safety of an invasive neural interface, a sensor array surgically implanted directly on top of the brain’s motor cortex and a decoder system that can take the brain signals recorded by the sensor array and translate that information into commands that can control a computer cursor or robotic arm.

“This trial has benefitted from nearly 40 years of fundamental basic science research that has helped us understand how the motor cortex and other parts of the human brain control movement,” says Leigh Hochberg, a neurologist and engineer at Brown University and the principal investigator and lead clinical investigator of BrainGate. “We’re working to develop and test new methods to enable neural control of an external device.”

The purpose of the current trial is to assess BrainGate’s safety. The trial has currently tracked eight people with tetraplegia due to diseases like amyotrophic lateral sclerosis (ALS), spinal cord injury, or stroke for at least one year of use.

“The general question being asked is whether this device is safe over one year post-implant,” says Hochberg. “Many people with tetraplegia are at an increased risk of illness due, for example, to skin breakdown from immobility, skin or deeper infections, pneumonia from the inability to clear oral secretions, urinary tract infections, and a host of other illnesses. Like any clinical trial, we keep track of these, even though those events are not related to the research. We also, of course, track any adverse events related to the device itself.”

BrainGate made headlines last year when one of the trial’s participants, a woman who had been paralyzed by stroke for more than 15 years, directed a robotic arm to pour a cup of coffee using by her thoughts alone. “Remember, depending on whom you ask there are 86 or 100 billion neurons in the brain. On a great day, BrainGate is recording from maybe 100 of them,” he says. “The fact that we can take that little sample of neural activity and turn it into control of a device that can assist with the activities of daily living is really exciting.”

But Eugene Alford, a certified plastic surgeon who works with Contreras-Vidal on the Rex Bionic Legs project, a non-invasive BMI that involves using an EEG cap to track brain signals to move the robotic legs, has concerns about the invasiveness of the BrainGate implants. “As a plastic surgeon, I remove a lot of non-human tissue from bodies. Any kind of implant that you put into the body that doesn’t have its own blood supply has a high risk of becoming infected,” he says. “The nice thing about Rex is that it uses a non-invasive brain mapping program based on an EEG electrode skull cap. It doesn’t require an invasive procedure.”

The EEG cap mapping has not proven to be quite as effective as the BrainGate type array—but Alford thinks the benefits still outweigh the risks. “It may be a little further off from where it needs to be than the more invasive interfaces,” says Alford. “But it’s worth the wait as far as I’m concerned.”

In the meantime

Hochberg is the first to say that, while the current results from the BrainGate study are promising, researchers cannot draw any definitive conclusions about the array’s safety, nor how effective it may be to use in a uncontrolled, non-laboratory setting. Yet, he remains optimistic.

“The dream, for us, is to one day reconnect the brain to the body in these patients in order to provide 24/7 control of a computer cursor or assistive device. I think that is an easily identifiable goal,” he says.

But Contreras-Vidal says that even if these technologies don’t live up to the hype, they still provide the scientific community invaluable information. “There are many challenges in this kind of work. But as we do it, we are gaining incredible knowledge about brain function and that also has great impact. If we understand, for example, the brain’s signature in certain areas and how that information is encoded or disrupted in disease, that will inform the development of new research and new approaches,” he says. “This is a win/win situation for everyone.”