This story is part of our December 2024 issue. To subscribe, click here.

In the summer of 2023, Casey Harrell had four tiny devices surgically implanted into the part of his brain that controls the muscles for speech. He’s in the prime of life: a 46-year-old father to his young daughter, a husband and a climate activist. But in 2020, he was diagnosed with amyotrophic lateral sclerosis, commonly known as ALS. Harrell can vocalize soft utterances, but he’s no longer able to produce intelligible speech.

The Blackrock Neurotech devices, known as microelectrode arrays, that UC Davis Health neurosurgeon Dr. David Brandman implanted in Harrell’s brain, record the electrical signals produced by his neurons as he attempts to speak.

Twenty-five days after the surgery, researchers from the UC Davis Neuroprosthetics Lab, co-directed by Brandman and Sergey Stavisky, a UC Davis neuroscientist, visited Harrell and his family at home. They connected him to the brain-computer interface they’re developing — a neuroprosthesis to restore speech — and calibrated the system. Within minutes, the words Harrell wanted to speak appeared on a monitor and were spoken aloud by a computer in a voice resembling his own using recordings from videos made before his diagnosis. It brought everyone in the room to tears.

“On day two of this use by Casey, he was talking to his daughter for the first time in her memory. That was so gratifying,” says Stavisky.

“I’d love to get to a point where I can meet a patient and say, ‘Ms. Smith, I’m sorry that you have ALS. This is a devastating diagnosis, and it’s going to impact you and everyone in your family. But I have a surgery that’s going to make sure that you can still speak with your daughters and let them know that you love them.’” Dr. David Brandman, Neurosurgeon, UC Davis Health

Stavisky and Brandman’s work is part of the BrainGate consortium, an ongoing clinical trial in its 20th year, studying the safety of devices that are implanted long term in people with paralysis. Researchers at various universities and hospitals throughout the country are focused on developing brain-computer interface technologies, known as BCIs, that safely restore mobility and speech lost to injury or disease. Harrell is a participant in the study. 

The majority of participants living with paralysis test BCIs designed to restore mobility, enabling them to control, for example, a robotic arm with their thoughts to perform tasks such as drinking from a cup without assistance.

Harrell is the second participant in the study to test a speech neuroprosthesis. After 84 data sessions collected over 32 weeks and more than 248 hours of self-paced conversations, the BCI developed by Stavisky, Brandman and their team accurately interprets what Harrell is trying to say roughly 97 percent of the time.

The results, which were published in The New England Journal of Medicine in August, significantly outperformed previously tested speech neuroprosthetics and smartphone applications that are designed to interpret a person’s voice. In the future, the technology could potentially help millions of people around the world restore basic functions of movement and communication and regain some control and independence over their bodies and environments.  
 

Trapped inside one’s body  

ALS is a progressive disease. “Little by little, the brain loses the ability to communicate with different muscles,” says Nancy Wakefield, managing director of care services for the ALS Association. “The losses come, in some cases slowly; in some cases quite rapidly.”

The disease can affect each person differently, explains Wakefield, and there’s no way to predict if it will selectively affect groups of muscles needed to eat, walk, write, talk and eventually breathe, or if it will attack them all. But it leaves a person’s cognitive abilities intact, stranding a flourishing mind inside a declining body. Research in Nature says most people diagnosed with ALS will “progressively lose the ability to speak.”

“He’s incredibly intelligent, very articulate, except he lives with paralysis. And so he has all these wonderful things to say, and he can’t say them easily because when he tries to speak through an expert interpreter, he says about roughly six correct words per minute,” says Brandman.

Expert interpreters are professionals who are trained to understand and facilitate communication for people with speech loss. But the average person speaks roughly 120-150 words per minute, making six correct words a minute insufficient to communicate well.

There are other devices on the market. Some are as simple as microphones to amplify speech or laser pointers to identify words or pictures on charts. There are apps that speak written text and computers that can track eye movement to select words on a screen and generate speech, as well as search the web, manage email and control things such as TVs and lights. But these can be exhausting or so slow that they can’t accommodate the rhythm of conversational speech, explains Wakefield. And some people may lose the ability to control their eye movement as the disease progresses.

Translating brain signals into speech

When a person formulates words they want to say, the brain sends specific signals to the muscles that control speech in order to produce the correct words. The brain breaks down words into units of sound called phonemes. For example, the word “dog” has three phonemes: d/o/g. Each one generates a unique electrical signal in the brain — like a fingerprint.

The microelectrode arrays implanted in Harrell’s brain record these signals and transmit them to a computer that’s connected to his head when the system is in use. Every 80 milliseconds, the BCI’s algorithm decodes his brain’s electrical impulses into phonemes, then words and sentences.  

Over several months following surgery, Stavisky, Brandman and several members of their team visited Harrell at home and collected data on the system’s ability to accurately decode what he wanted to say. Snippets that appeared onscreen from a condensed video recording of Harrell’s conversational speech convey the capability — and the value — of the BCI’s technology. These were recorded during Harrell’s 30th session as he sat in his power-lift wheelchair next to an undecorated Christmas tree. This is how his messages — produced without capitalization and punctuation — are displayed on the computer:

one of the things that people with my disease suffer from is isolation and depression because they do not feel like they matter anymore and something like this technology will help bring people back into life and into society
 
that really cannot be understated how important that is
 
so I think about this from my personal perspective and also the perspective of people like me who might not be as lucky but deserve the same treatment
 
i hope that we are very close to the time when everyone who is in a position like me has the same option to have this device as i do

As a participant in the clinical study, Harrell uses the BCI up to 12 hours a day in his home to communicate with family, friends and coworkers. He still works as a climate activist.  

Harrell speaks of the people in his position — those who have lost basic functions due to injury or disease. So many people are affected worldwide. Large percentages of people living with Huntington’s disease, multiple sclerosis, stroke, traumatic brain injury and Parkinson’s disease also experience dysarthria, a neuromotor disorder that affects the muscles required to produce speech, according to the American Speech-Hearing-Language Association. And roughly 5.4 million — or 1 in 50 people in the U.S. — live with some form of paralysis, according to research by the Christopher and Dana Reeve Foundation.

“I’d love to get to a point where I can meet a patient and say, ‘Ms. Smith, I’m sorry that you have ALS. This is a devastating diagnosis, and it’s going to impact you and everyone in your family. But I have a surgery that’s going to make sure that you can still speak with your daughters and let them know that you love them,’” says Brandman. We’re not there, he says, but realistically he thinks we will be in five to 10 years.

Funding the future

The way that’ll happen, says Brandman, is companies will work with academic laboratories, such as the UC Davis Neuroprosthetics Lab, to build devices that restore function through industry-sponsored studies. Company-led clinical trials often hire medical doctors who work in university settings as a study’s principal investigator. The company funds the study and retains ownership of the research, explains Brandman.  

Investing in academic research isn’t always attractive to corporate interests. The research can take years, and the work is often exploratory with lots of risk and uncertainty, which can make it difficult to align investments with a company’s short-term goals to market products and achieve immediate returns. 

But now that Stavisky, Brandman and their team have provided a demonstration of the technology’s potential for meaningful communication, it’s led to multiple conversations with venture capitalists who read their paper in the New England Journal of Medicine and are interested in investing in this space, says Stavisky. 

“I hope that this means that startup teams that want to build communication neuroprostheses have more tail-wind,” he writes in an email. He says he sees the research ecosystem working exactly as it should. “An academic clinical trial shows for the first time that something is possible, even if the system is bulky and not in the form factor that we’d ideally see. This de-risks it and leads to investment in companies that can take this proof-of-feasibility and refine it and then scale it.”

Funding for the UC Davis Neuroprosthetics Lab includes more than $5.7 million in grants awarded from the U.S. Department of Defense Congressionally Directed Medical Research Program on ALS, the National Institutes of Health Director’s Innovator Award, the Pershing Square Foundation Maximizing Innovation in Neuroscience Discovery Prize and the ALS Association’s Assistive Technology Grant. Additional support has come from various foundations, as well as startup funding and seed grants from UC Davis.

Relatively uncommon in the academic lab space, Stavisky and Brandman co-direct their lab, a model Stavisky saw work successfully at Stanford University, where he trained at its BCI lab. 

“If you look at industry, think of a startup, there’s typically co-founders. It is very hard to do everything alone. And I think that’s actually one of the unfortunate parts of the way academia is set up, is that a professor runs a lab alone, and they need to be good at absolutely everything or at least tolerable at it,” says Stavisky. 

As a pair, they’re able to divide the workload, leverage their strengths and advance their research faster. Brandman leads the medical side and the clinical trial, and Stavisky leads the basic science. Together they work on the engineering. 

In the future, the team plans to expand their research to restoring memory and language, for example, aphasia — a neurological disorder that makes it difficult to understand or use language. 

There are roughly 2 million people in the United States alone with aphasia, according to The National Aphasia Association. Brandman speaks of children with cerebral palsy who don’t have the mechanics of speech because they never developed properly. Brandman asks, can we help them restore function they’ve never really had?

But right now, they’re specifically trying to understand speech formation in the brain and help people who want to talk but can’t: people who don’t have problems with language or understanding others. It’s just the mechanics of speech that aren’t working because of an injury or disease, explains Brandman.  

“That’s arguably a fundamental human right, that we ought to be able to say the things we want to say,” he says.  

Stay up to date on business in the Capital Region: Subscribe to the Comstock’s newsletter today.