Enlarge /. Human brain, motherboards, chips and artificial intelligence concept, as well as neural technology and brain computer interfaces.
The hard part of connecting a sticky, thinking brain to a cold computer that is one and zero is getting information through your thick skull – or mine, or anyone. After all, a skull is about safely separating a brain from it (beckoning your hands).
So if that brain is not your brain, the only way to tell what is going on inside it is by inference. People make very educated guesses based on what that brain is telling a body – for example, when the body is making sounds that you can understand (that's language) or that move in recognizable ways. This is a problem for people trying to understand how the brain works, and an even bigger problem for people who cannot move or speak due to injury or illness. Sophisticated imaging technologies like functional magnetic resonance can give you some pointers. But it would be great to have something more direct. For decades, technologists have tried to get the brain to connect to computer keyboards or robotic arms to associate meat with silicon.
On Wednesday, a team of scientists and engineers showed results of a promising new approach. Electrodes are attached to a flexible, resilient tube, a so-called stent, and passed through a blood vessel that leads to the brain. When testing two people, the researchers literally opted for the carotid artery and ran a stent-tipped wire through the vein in the neck and then into a vessel near the brain's primary motor cortex, where they burst the source. The electrodes snuggled into the vessel wall and began to recognize when people's brains signaled their intention to move – and sent these signals wirelessly to a computer via an infrared transmitter that was surgically inserted into the subjects' chest. In an article published in the Journal of NeuroInterventional Surgery, the Australian and US researchers describe how two people with paralysis due to amyotrophic lateral sclerosis (better known as Lou Gehrig's disease) used such a device to send texts and play around online through brain control alone.
“The self-expanding stent technology has been well demonstrated in both cardiac and neurological applications for the treatment of other diseases. We just use this feature and put electrodes on the stent, ”said Thomas Oxley, an interventional neurologist and CEO of Synchron, the company that wants to commercialize the technology. "It's fully implantable. Patients will go home in a few days. And it's plug and play."
It took a workout once the subjects got home. The electrode-studded stent could pick up signals from the brain, but machine learning algorithms need to figure out what those signals – incomplete reflections of a mind at work, even under ideal conditions – actually represent. After a few weeks of work, both patients were able to use an eye tracker to move a cursor and then click the implant with a thought. It doesn't sound like much, but that was enough for both of you to text messages, shop online, and otherwise engage in everyday digital activities.
The Food and Drug Administration has not yet approved what Oxley is calling a "stentrode" for widespread use, and the company is still looking for more testing, but these preliminary results suggest it is a working brain-computer Interface acts. The received signal is not full of information. At the moment, the Stentrode only records one piece of information – either a telepathic mouse click or the absence of this click. But for some applications this may be enough. "There's been a lot of talk about data and channels, and what really matters is, have you delivered a life-changing product to the patient?" Oxley says. "Only with a handful of expenses restored for the patient they are in control of can they control Windows 10."
Much more ambitious brain-computer interfaces and neural prostheses have been in the news lately. Last month, Elon Musk's Neuralink company demonstrated a wireless BCI with more than a thousand flexible electrodes that can be inserted directly into a brain by a specialized robotic surgeon. (The company has shown only short-term use in pigs so far.) Electrode insertion is difficult; While it is true that brain surgery is not a rocket science, it does have risks whether the surgeon is a robot or not. Even flexible, thin electrodes like those found by Neuralink are invasive enough that the brain tries to defend itself against them, coating them with glial cells, which reduce their ability to conduct the electrical impulses they want. And while implanted electrodes like those of the more commonly used "Utah array" can receive clear signals from individual neurons, understanding the meaning of these signals is still underway in science. Plus, the brain sloshes around like jelly in a donut; Permanently installed electrodes can damage it. But get it right and you can do more than just brain research. Caged patients with ALS have used them as successful brain-computer interfaces even though they require training, maintenance, surgery, etc.
Meanwhile, electrodes placed directly on the scalp can pick up brain waves – electroencephalograms or EEGs – but lack the spatial details of implanted electrodes. Neuroscientists have a very rough idea of what part of the brain is doing what, but the more you know which neurons are firing, the better you can tell what they're firing at.
A recent innovation, electrocorticography, places a mesh of electrodes directly on the surface of the brain. Combined with the intelligent spectral processing of the signals picked up by these electrodes, the EKG is good enough to convert into text or even speech the action in the part of the motor cortex that controls the lips, jaw, and tongue. And there are other approaches. CTRL Labs, which Facebook bought for maybe $ 1 billion in 2019, are trying to get motor signals from neurons in the wrist. The kernel uses functional near-infrared spectroscopy on the head to capture brain activity.
The stentrode from Oxley and his colleagues, if it continues to show good results, fits somewhere between implanted electrodes and EEG. The inventors hope that it will be closer to the first than the second. But it's still early. "The core technology and idea are super cool, but considering where they're accessing the signals from, I'd expect this to be a relatively poor fidelity signal compared to other brain-machine interface strategies" says Vikash Gilja. He heads the Translational Neural Engineering Laboratory at UC San Diego. "At least we know that a high-density EKG recording from the surface of the brain can convey information beyond what is shown in this article."
One potential problem: tissue conducts electrical impulses, but the electrodes in the stent pick up signals from the brain through the cells of the blood vessel. That lowers the signal content. "If we were to compare these cortical surface recordings to Utah array experiments – most of the clinical experience with implanted electrodes – I would say the type of recording on the EKG is a rate limiter," says Gilja. (Just for the sake of transparency, I should point out that Gilja has done paid work with BCI companies like Neuralink that Synchron could theoretically compete with one day.)
So it might not be good enough for neuroscience, but it can be useful for a person with paralysis who wants a low-maintenance BCI that doesn't require drilling through the skull. "There is a tradeoff between how invasive you want to be and what level you want to gather information," says Andrew Pruszynski, a neuroscientist at Western University in Canada. “This tries to get to the center and insert a catheter close to the neural activity. It's obviously invasive, but certainly not as invasive as inserting electrodes into the brain. "
And there is more work to be done. Oxley's team hopes to expand its study to include more human subjects. They will look for possible side effects, such as the possibility of the stent contributing to stroke (although this seems less likely because it embeds itself in the vessel walls, a process called endothelialization). You could find better locations for the stent in blood vessels that border other brain regions of interest. Somewhere within 2 millimeters of a vessel large enough to hold the stentrode is fair game, Oxley says. The software could do better at figuring out what the brain actually means when it sheds its electrical bells and whistles, and some of their tests suggest that the system could capture more information details – like what specific muscles the users were trying to contract . This could lead to more useful prosthetics or control of devices outside of Windows 10. "The motor system is currently going to provide therapy for people who are paralyzed," says Oxley. "But when we start looking at other areas of the brain, you start to see how technology will open up the processing power of the brain." It's hard to predict what could happen if scientists actually figure out how to get it into someone's head.
This story originally appeared on wired.com.