TL; DR
A UC Davis BCI implant allows an ALS patient to speak independently for 3,800+ hours with 99% accuracy over two years, allowing him to work full-time.
An ALS patient has used a brain implant to speak independently for more than 3,800 hours over the past two years.produces about 2 million words at an average rate of 56 words per minute. The study, published Monday in the journal Nature Medicine by researchers at the University of California, Davis, represents the longest continuous demonstration that a brain-computer interface can function as a practical everyday communication tool outside the laboratory. Participant Casey Harrell, 47, used the system to return to full-time work as an environmental activist.
The implant consists of an array of four microelectrodes placed in Harrell’s left precentral gyrus, a brain region that coordinates speech, recording activity from 256 cortical electrodes. Built on a software platform called BRAND, developed by UC Davis doctoral student Nicholas Card, machine learning algorithms convert that neural activity into English phonemes, then map those phonemes into words and sentences. The system reads the decoded text into a synthesized version of Harrell’s pre-ALS voice.
In a test run with a 125,000-word dictionary, the system achieved over 99% word accuracy. In everyday use outside the lab, Harrell correctly or mostly correctly named 92% of the sentences. During the study period, he delivered more than 183,000 sentences.
💜 of EU technology
The latest rumblings from the EU tech scene, a story from our wise founder Boris and some questionable AI art. Free in your inbox every week. Register now!
“The bottom line for me is that it allows for everyday communication for a guy who wants to talk but can’t.Neurosurgeon David Brandman, who implanted the device in 2023 and led the study, told The Register.Despite her paralysis, she returned to her full-time job and had meaningful conversations with her daughter, who had never heard her voice.“
The importance of research lies not only in accuracy, but also in independence. Previous BCI systems required researchers to be at the patient’s home or the patient to visit the laboratory when the device was used. Harrell’s system is administered by a home care team without the need for any researcher support.
According to the study’s timeline, he used it for an average of more than five hours a day.
He is part of the UC Davis team BrainGate, a consortium of universities and the US Department of Veterans Affairs, is developing brain-computer interfaces for speech restoration, computer control and movement restoration. The apparatus itself is not custom built using existing microelectrode arrays manufactured by Blackrock Neurotech. The breakthrough is in the software, specifically the BRAND platform’s machine learning algorithms that decode attempted speech from neural signals in real-time.
Brandman compared the current state of BCI technology to early pacemakers in the 1950s, which required external wiring to large batteries or wall power. Seventy years later, pacemakers are implanted in outpatient procedures. “We are in the early stages of this kind of technology,The fireman said.
Harrell is still attached to external computers, but the UC Davis team’s AI development, along with hardware miniaturization work at companies like Neuralink, Synchron, and Paradromics, point to a future where setups will be less difficult.
The competitive environment at BCI is accelerating. Neuralink has implanted devices in at least 21 patients under research protocols, but does not have commercial approval. China approved the first commercially available invasive BCI earlier this year.
Other approaches to restoring speech for people with ALS use AI speech conversion not brain implants, but these techniques require the patient to retain some voice ability.
What distinguishes the UC Davis work is that it demonstrates that BCI can move from a laboratory experiment to a sustainable, practical everyday tool. According to principal investigator Sergey Stavisky, the 3,800 hours of brain recordings also constitute the largest single-neuron-resolution data set ever collected, which will inform future improvements in decoding algorithms.
The system remains an investigational device limited by federal law to research use and tested on one patient. It is not yet known whether the results generalize to other ALS patients or people with different neurological diseases. Scaling the technology from a clinical trial to a designated medical device requires regulatory approval, device miniaturization, and cost reduction, which can take years.
“I really want to not be unique or special because that will mean I won’t get the disease anymore or that anyone else who has the same disease as me can buy them this prescription,Harrell said through the BCI system.
