For nearly three years, Casey Harrell has lived with a device embedded in his brain that has transformed how he communicates. Paralyzed by amyotrophic lateral sclerosis (ALS), a progressive disease that gradually removes control over muscles, Harrell has used a brain-computer interface (BCI) to turn his thoughts into spoken words, giving him a new way to interact with the world.
The technology, developed by researchers at the University of California, Davis, has allowed Harrell to communicate with friends and family, read to his young daughter, browse the internet, and continue his work. His extensive use of the system has now positioned him as what researchers describe as the first “power user” of a speech BCI.
The findings, reported by MIT Technology Review and published in the journal Nature Medicine, show that Harrell used the device for more than 3,800 hours at home during the first 22.6 months after implantation without researchers being physically present. The milestone represents one of the longest periods of independent daily use reported for a speech-focused brain-computer interface.
“Living with a disease like ALS, you are supposed to have diminished dreams. I do not,” Harrell told MIT Technology Review. He said the ability to regain communication and independence through the device has changed his daily life.
Harrell first began working with David Brandman, an associate professor of neurological surgery at the University of California, Davis, and his research team after he was diagnosed with ALS. By the time he joined the trial, he required assistance to control his wheelchair and complete basic tasks such as dressing and eating. His speech had also become difficult to understand.
When researchers asked whether he wanted to participate in testing a brain implant designed to help restore communication, Harrell agreed. At the time, he was 45 years old and wanted to contribute to the development of a technology that could help people with severe paralysis communicate.
In July 2023, surgeons implanted four arrays containing 64 electrodes each into Harrell’s brain. The electrodes were positioned in areas connected to speech production and linked through two external connection points on his skull, allowing brain signals to be transmitted to a computer.
The research team developed algorithms capable of interpreting activity from the speech motor cortex, the region of the brain responsible for controlling movements involved in speaking. Instead of directly reading thoughts, the system identifies patterns associated with the sounds people produce when forming words.
“There are 39 phonemes that make up all the sounds in the [American] English language,” Nicholas Card, a neuroengineer at UC Davis and member of the research team, explained. By mapping brain activity linked to these sounds, researchers created a personalized speech decoder that converts neural signals into phonemes and then into words.
The system began working approximately one month after Harrell’s surgery. On the first day of testing, the speech decoder allowed him to communicate using a 50-word vocabulary with 99.6% accuracy. Over time, the vocabulary expanded to approximately 125,000 words while maintaining 97.5% accuracy.
The long-term performance of the implant was initially uncertain. Researchers have noted that brain implants can face challenges over time, including the development of scar tissue around electrodes that may interfere with signal detection. However, Harrell’s device has continued to function, allowing scientists to study how the technology performs outside controlled research settings.
A significant development has been the increased independence of the system. In the early stages, members of the research team had to visit Harrell’s home to connect and disconnect him from the device whenever he wanted to use it. The system has since been automated further, allowing his care partner to handle the setup.
“He’ll wake up, get plugged in, and just get going,” said Sergey Stavisky, a neuroengineer at UC Davis and member of the team.
Researchers say this transition from laboratory use to everyday life is essential for evaluating whether BCIs can become practical tools for people with disabilities. Mariska Vansteesel, a BCI researcher at Utrecht Medical Center who was not involved in the trial, said technologies intended for patients need to be tested in real-world conditions to determine whether they are usable and valuable without constant support from researchers.
The system has also continued to evolve based on Harrell’s needs. Researchers have improved its accuracy to about 99%, added cursor control capabilities, and enabled him to use his computer for sending messages, writing emails, browsing the web, and managing his work as an environmental activist.
Additional software features were created in response to Harrell’s personal requests. The system now includes a privacy mode that automatically deletes decoded text and a profanity filter that can be used while he communicates with his daughter.
“We have been able to add on to the software side of the device … improving the accuracy and adding more bells and whistles to enable me to be more independent when using the device,” Harrell said.
Despite the progress, researchers caution that Harrell’s results may not represent what every person with ALS will experience. Long-term success can vary depending on individual health conditions, and other cases have shown that implanted BCIs may eventually lose effectiveness as neurological diseases progress.
Some researchers also note that invasive brain surgery remains a barrier for many potential users. Jane Huggins, a University of Michigan researcher developing noninvasive BCIs who was not involved in the trial, said many people with progressive conditions may be reluctant to undergo surgical procedures despite the potential benefits.
For Harrell, the technology has provided practical changes in his daily life. He said the device has allowed him to continue working, maintain financial support for his family, reconnect with friends, and participate more actively in raising his daughter.
Researchers are continuing to develop the system, with the goal of eventually creating a “brain-to-voice” technology that could convert brain activity directly into a natural-sounding voice with emotional expression, including changes in tone and inflection.
After years of using the device, Harrell said the results have exceeded what he expected when he first joined the trial. The ongoing research continues as scientists work to improve communication technologies for people living with severe neurological conditions.
