A research team led by Frank Willett, PhD, assistant professor of neurosurgery at Stanford University, has developed a brain-computer interface (BCI) that can decode inner speech in people with paralysis. The technology aims to help patients who are unable to speak clearly due to severe motor impairments.
The BCI uses microelectrode arrays implanted in the surface layer of the brain’s motor cortex to record neural activity patterns. These signals are then processed by computer algorithms trained through machine learning to recognize patterns associated with phonemes—the basic units of speech—and assemble them into sentences.
Frank Willett explained the potential impact of decoding inner speech for those with paralysis: “Inner speech (also called ‘inner monologue’ or self-talk) is the imagination of speech in your mind – imagining the sounds of speech, the feeling of speaking, or both. We wanted to know whether a BCI could work based only on neural activity evoked by imagined speech, as opposed to attempts to physically produce speech. For people with paralysis, attempting to speak can be slow and fatiguing, and if the paralysis is partial, it can produce distracting sounds and breath control difficulties.”
In a recent study published August 14 in Cell and conducted with colleagues from several institutions including Emory University and Harvard Medical School, researchers worked with four participants who had microelectrode arrays placed in their brains. They found that inner speech produced clear patterns of activity similar—though smaller—to those generated by attempted physical speech.
Willett noted: “We studied four people with severe speech and motor impairments who had microelectrode arrays placed in motor areas of their brain. We found that inner speech evoked clear and robust patterns of activity in these brain regions. These patterns appeared to be a similar, but smaller, version of the activity patterns evoked by attempted speech. We found that we could decode these signals well enough to demonstrate a proof of principle, although still not as well as we could with attempted speech. This gives us hope that future systems could restore fluent, rapid, and comfortable speech to people with paralysis via inner speech alone.”
The researchers also addressed privacy concerns related to accidentally decoding thoughts not intended for communication. According to Willett: “The existence of inner speech in motor regions of the brain raises the possibility that it could accidentally ‘leak out’; in other words, a BCI could end up decoding something the user intended only to think, not to say aloud… Nevertheless, we’re proactively addressing the possibility of accidental inner speech decoding, and we’ve come up with several promising solutions.”
One solution involves training BCIs designed for physical-speech decoding to ignore inner-speech signals more effectively; another introduces a password-protection system for next-generation devices so no inner thoughts are decoded unless a specific phrase is imagined first.
“For current-generation BCIs… we demonstrated in our study a new way to train the BCI to more effectively ignore inner speech… For next-generation BCIs… we demonstrated a password-protection system that prevents any inner speech from being decoded unless the user first imagines the password (for example… ‘Orange you glad I didn’t say banana’). Both methods were extremely effective at preventing unintended inner-speech from leaking out,” Willett said.
Looking ahead, improved hardware—such as fully implantable wireless devices—is expected within several years. Researchers are also exploring other areas beyond the motor cortex for potentially richer information about imagined language processing.
Willett summarized future directions: “Improved hardware will enable more neurons to be recorded and will be fully implantable and wireless… To improve accuracy… we are also interested in exploring brain regions outside of the motor cortex…”
This research was originally published by Stanford Medicine.
