Brain device may one day help people with ALS communicate
Fluent’s interface is designed to turn attempted speech into words
Written by |
- A less invasive brain-computer interface is being developed to help people with impaired speech, including ALS, communicate.
- Fluent’s device would sit under the scalp but outside the skull, which the company says could improve safety and access.
- The system aims to turn attempted-speech brain signals into text or audio, with clinical studies planned later this year.
A University of Melbourne spinout company is developing a less invasive brain-computer interface that could one day allow people with impaired speech, including those with amyotrophic lateral sclerosis (ALS), to communicate by attempting to say words.
Unlike many current brain-computer interface prototypes, which can involve invasive surgery with electrodes implanted inside the skull, Fluent’s low-risk interface is designed to sit beneath the scalp but outside the skull. The company says this less invasive approach has a better safety profile than a cochlear implant, a surgically implanted hearing device, and could make the technology more widely accessible.
“Until now, this technological capability was only thought possible using highly invasive electrodes implanted inside the skull,” Tim Mahoney, PhD, biomedical engineer and Fluent’s CEO and co-founder, said in a university news release. “With a safety profile that’s even better than a routine cochlear implant, the technology will be more accessible to the broader population.”
Brain signals may aid communication
In ALS, the nerve cells that control voluntary movement, called motor neurons, gradually die, leading to worsening muscle weakness and increasing difficulty with movement, speaking, swallowing, and breathing.
Current communication tools, including communication boards and eye-tracking systems, allow many people with ALS to spell out words and sentences by selecting letters on a screen through touch or eye movements. However, these systems can become increasingly difficult to use as muscle weakness progresses.
Brain-computer interfaces aim to bypass those physical limitations by translating brain activity directly into commands for a computer. Fluent’s device is designed to record electrical signals from above the motor cortex, the region of the brain that controls the muscles used for speech.
Even when a person can no longer speak, the brain may still generate distinct patterns of activity when they attempt to say words. Researchers are training a machine learning model to recognize those patterns and convert them into text or audio, with the goal of allowing people to communicate without speaking or pressing a button.
“If you think of electrical signals as QR codes: when a person speaks, every individual mouth and jaw movement produces different ‘QR codes’ in their motor cortex,” Mahoney said. “These ‘QR codes’ also occur when a person with impaired speech is attempting to speak. Our device can capture these codes in a sequence, which tells us what someone is trying to say.”
What sets Fluent’s technology apart is where those brain signals would be recorded. Current prototypes often involve invasive surgery using electrodes implanted inside the skull. Fluent’s device instead is designed to sit beneath the scalp but outside the skull, avoiding direct contact with the brain while still aiming to capture high-quality electrical signals.
Clinical studies planned for later this year
During his doctoral research, Mahoney showed that the quality of electrical signals captured with the device was comparable whether it was placed beneath or outside the scalp, allowing the team to develop the technology without the need for complex surgery.
The technology has already undergone preliminary human testing with participants at the Aikenhead Centre for Medical Discovery in St Vincent’s Hospital Melbourne.
“Participants had 144 electrodes placed on their scalp and the electrodes recorded brain activity occurring in their motor cortex while they spoke, mimed and imagined saying different phrases,” Mahoney said.
Using what the team describes as the largest English-language dataset of its kind, together with an even larger dataset from collaborators in Japan, the researchers developed a machine learning model that identified the correct phrase from a pool of 128 options with 96% accuracy.
Fluent has raised more than $2 million to advance the interface toward clinical studies of its insertable electrodes, which are scheduled to begin later this year.
The technology could ultimately help people with ALS and other neurological diseases regain a greater degree of independence, according to Mark Cassidy, PhD, deputy vice-chancellor (research) at the University of Melbourne.
Leave a comment
Fill in the required fields to post. Your email address will not be published.