The often-secretive brain-implants industry is clearly having a moment.
On the one hand, a group of Swiss and French doctors confirmed they had enabled a Dutch paraplegic man to walk again simply by asking the patient to think about it and using artificial intelligence to read brain signals sent from wireless chips implanted in his spine and brain.
And then Elon Musk’s controversial startup Neuralink finally received US regulatory approval from the Food and Drug Administration (FDA) to begin testing on human patients, after multiple failed attempts.
Mr Musk has previously painted an exciting picture of the future, whereby Neurolink’s brain implants will one day allow its customers to impose their will on the world using just thoughts. Unsurprisingly, some people are falling over themselves to be the guinea pigs.
What brings these two announcements together is that they both involve invasive surgical procedures. However, many experts believe this is unnecessary and foolhardy.
“It’s crazy that you need to drill into the skull to solve the health challenges we do today,” says Dr Nataliya Kosmyna, a research scientist at MIT’s Media Lab.
She also point out that everyone is getting carried away with the promise of what Neurolink might – or might not – deliver, given the limited information it has shared so far.
“FDA approval doesn’t mean the system will work — it can be taken away. It doesn’t mean FDA confirms it’s the way to go, it just offers you the possibility to test on humans.”
Scientists have been researching brain-computer interfaces (BCI) for decades. This covers a wide array of technologies that might seem like the stuff of sci-fi, except that they already exist – and many of these ideas do not require a surgeon’s scalpel. Indeed, there are three main types: exoskeletons, wearable devices, and implantable chips.
For instance, it’s already possible to beam information from your mind to a machine telepathically. And these techniques are not mere frivolous fun – they open a raft of possibilities to help people with medical conditions such as epilepsy.
In her Neurafutures website, Dr Komyna has calculated the likelihood of the futuristic technologies showcased in movies, video games, TV series, or books becoming viable.
So are brain implants about to go mainstream, or is this a geeky pipe-dream? The Standard asks a cluster of world-class experts for the lowdown here. Buckle-up for a wild ride into the realm of the brain machines…
Robot control
Ever wanted to remotely control a robot with your mind? Well, that’s already possible today, says Dr Kosmyna.
She has built a system that lets her brain control Boston Dynamics’ robot dog Spot, using glasses, a mobile app, and two iPhones.
Sensors on the glasses’ frame sense EEG brainwaves and eye movements corresponding to specific requests. When Spot asks if it should go to the kitchen, Dr Kosmyna thinks of the keyword “yes”. The system sends her answer to the robot, which obeys.
Neurotechnology has already had a significant impact on patients with otherwise intractable conditions
MIT has tested controlling other objects, too. “We’ve shown in lab tests that you can use Philips Hue [smart lights] to turn on your lights with your mind, and you can think of the kettle and it will turn on,” says Dr Kosmyna.
“All of this stuff is achieveable and possible but it’s expensive to even have them in the lab... some of the systems cost $10,000 (£8,020).”
Since at least 2013, we’ve also had exoskeletons. Paralysed patients can train their brains to work with the large, often clunky robotic suits and issue commands for them to walk and move, while the suit supports the body.
The technology can also be used by people in professions where a lot of heavy lifting is needed, such as in factories or the military.
One famous example is the 2014 Fifa World Cup opening ceremony, where paralysed Brazilian patients were able to kick the ball — the work of brain-computer interfaces pioneer Dr Miguel Nicolelis.
Telepathy
In the 2013 film Pacific Rim, giant robots called Jaegers are controlled by two pilots whose minds are joined by a telepathic mental link.
This sounds like a glimpse of the future but according to Dr Kosmyna, several types of brain-to-brain communication are already possible.
A 2019 study by the University of Washington and Carnegie Mellon researchers showed that three people could work together to win a Tetris-like game.
Two people transmitted information remotely on whether to rotate a Tetris block. The third person received a flash of light in their left or right eye, which told them to rotate the block clockwise or anti-clockwise.
However, before we get too excited, there are currently limitations.
I’ve always dreamed of being able to simply think of words and have my thoughts appear on a PC screen immediately. Unfortunately, this isn’t quite possible yet, says Dr Kosmyna.
“Word for word is not possible [right now], but we can detect keywords.”
MIT is trialling a system called Brain Switch, where a quadriplegic patient can communicate their basic needs to a carer using a pair of glasses wearable device. It works even for people who are totally paralysed, with locked-in syndrome, and those who cannot move their eyes.
Sensors on the glasses frame detect EEG brainwaves and eye movements corresponding to specific requests, such as “yes”, “no”, “turn me”, “put on music”, or “I need eyedrops”.
“Ninety-three per cent of Americans refuse intubation because they think it really is the end, but the brain is still functioning with [a progressive neurodegenerative disease like Lou Gehrig’s disease, also known as] ALS,” says Dr Kosmyna.
“We give the families iPods. It's an iOS app, they can bring it to the hospital and nurses can use it to communicate with the patient.”
While the system cannot detect emotions, family members can request custom emotive words like “happy” or “sad”.
MIT has shipped more than 100 Brain Switch devices since 2021 globally. Families of patients are welcome to apply for a device, which collects brain data to further train the system.
Brain stimulation and data collection
Brain control is a popular trope in science fiction. In Star Trek: The Next Generation (1992), Worf is paralysed and doctors embed spinal implants that transmit neural signals to help him walk again. The show raises the question of research data collection being prioritised over patients’ lives.
Numerous startups have been working on chips and electrodes that are either embedded deep in the brain to stimulate tissue — to treat conditions like Parkinson’s disease or epilepsy — or attached to the brain stem to gather data about brain signals.
The families I work with have this high expectation because they see it in the media, so they think it’s about to be imminently available, and then research scientists bring them back to ground
For these purposes, a surgeon typically has to remove a piece of your skull or drill small holes.
In paralysed patients, parts of their nervous systems are damaged, meaning nerve cells are unable to carry messages from the brain to muscles.
With the Dutch paraplegic man who was able to finally walk again, researchers from Swiss Federal Institute of Technology (EPFL) and Lausanne University implanted chips in his brain and spine, creating a “digital bridge”.
The disc-shaped implants replaced damaged nerves, sending wireless messages from the brain to the chip in the spine. That chip then passed the signal on to healthy nerve cells that could understand the command.
Mr Musk’s Neuralink is also developing a round chip. According to Dorian Haci, a translational researcher at Imperial College’s Next Generation Neural Interfaces (NGNI) Lab, Neuralink is ”inventing the wheel” by designing a brain-computer interface device “from scratch” and a novel surgical robot for implanting the device into the brain.
We are building surgery simulations for faster iteration and better test coverage. Join us to help expand this capability🦾 #techtuesday pic.twitter.com/JHFM5HersL
— Neuralink (@neuralink) March 21, 2023
“After they have implanted all the electrodes, they cover the hole in the skull with the device itself. They have screws to hold it in place, then they put the scalp skin back on top,” he explains to The Standard.
Both Dr Kosmyna and Dr Haci agree that Mr Musk’s efforts have shone a welcome light on the brain-computer interfaces industry. While Neuralink’s plans originally sounded sketchy, they say the latest demonstrations by more seasoned researchers are much more in line with science.
Should we drill into the skull for brain implants?
But should we insert brain implants at all? The Swiss-French project required doctors to cut circular 5cm-wide holes in each side of the Dutch patient’s skull. Neuralink drills holes, too, but they are a few millimetres wide.
“Neurotechnology has already had a significant impact on patients with otherwise intractable conditions; for example, cochlear implants have now restored functional hearing to an estimated one million patients worldwide,” Dr Andy Greenfield, a member of the Regulatory Horizons Council, an independent body set up by the Government, told The Standard.
Cochlear implants are surgically placed inside the ear by making a hole in the skull.
Imperial’s Dr Haci has co-founded MintNeuro, a start-up that is developing brain-implant technology to better manage and treat neurological conditions.
He is in favour of invasive implants for medical purposes: “Whatever technology we develop in the future, if we place it closer to the neurons in the brain, it will allow for much better understanding of what's going on in the brain, and a much more precise and accurate intervention.”
The problem with invasive implants in the brain is the duration of the recordings — if the brain signals extinguish over the months or years after surgery, the technology is not going to be helpful
Dr Kosmyna is not keen on widespread use of invasive implants. Non-invasive wearables or minimally invasive implants inserted up the nose or in your teeth, which are still near the brain, have the most promise to help people if you need to collect data on brain activity, she says.
Prof Grégoire Courtine of EPFL, who led the Swiss-French project, disagrees: “Our solution is an optimal trade-off between long-term stability versus sufficiently invasive to be easy to use rapidly — current non-invasive recording methods are not realistic for applications in real life.”
The troubling ethics of brain implants
Experts agree the BCI industry is a “wild, wild west” that sorely needs regulation.
Whenever new research about helping paralysed people walk goes viral, patients and their families often end up hugely disappointed.
“The families I work with have this high expectation because they see it in the media, so they think it's about to be imminently available, and then research scientists bring them back to ground,” says Dr Kosmyna.
“They say, ‘Oh, but we saw this paper’. And I have to say, ‘This is not coming any time soon.’”
And what happens if you have a brain implant, it improves your life, but then the company goes out of business and it has to be removed?
This sad scenario happened to a woman in Australia who was in a trial by failed US startup Neurovista. The implant helped treat Rita Leggett’s epilepsy, preventing seizures and making her feel like she could do anything. She is now severely mourning the loss of the implant.
German and Australian ethicists who studied Ms Leggett’s case in a new paper concluded that Neurovista impacted her neural integrity by essentially “creating a new person”. They recommend further discussions on human rights relating to brain implants.
“I cannot stress how important this is — all of these companies need money. They're hopefully helping people, they're pushing research and science, but they are for profit. Some of them will go out of business, some of their research will remain only research,” Dr Kosmyna says.
Another question is whether you can take an old implant out and replace it with a newer model.
“The problem with invasive implants in the brain is the duration of the recordings — if the brain signals extinguish over the months or years after surgery, the technology is not going to be helpful,” says Prof Courtine.
Computer scientists have grave concerns about how patients’ data will be used, particularly with the rising advent of generative artificial intelligence. Dr Kosmyna imagines a nightmare-inducing scenario where tech giants could one day show ads on our eyeballs.
The Regulatory Horizons Council made 14 recommendations in November to the Government about a number of issues, including potential under-regulation of non-medical neurotechnologies, and the collection and processing of data relating to the brain and other parts of the nervous system.
A spokeswoman for the Department of Health and Social Care told The Standard that it is currently reviewing the report’s recommendations.
“Recording data from the brain is among the most private information you can have on a person,” says Dr Haci. His research group is encrypting brain signals sent by the implants so they can’t be hacked.
“We need to consider decades in advance of how these technologies will be integrated into our world and also ask the right questions about ethics and regulation when it comes to data coming from the brain.”
