Brain-computer interface helps patient speak in real time

Paradromics’ investigational Connexus brain-computer interface (BCI) enabled a woman with a progressive motor neuron disease to generate her own words in real time, answer open-ended questions, and communicate with her family using synthesized speech.

The woman, who had a severe impairment of her ability to speak, was the first participant in the FDA-approved Connect-One Early Feasibility Study (NCT07357428). The BCI allowed her to choose what she wanted to say, marking a step toward restoring more natural communication for people who have lost the ability to speak, according to a press release.

The study is evaluating whether the device can safely restore communication in people with severe motor impairments caused by conditions such as amyotrophic lateral sclerosis (ALS). It is planned to enroll participants, ages 22 to 75 years, at three U.S. sites.

“When we implanted Connexus in June, we crossed the threshold into long-term human use. We have now crossed another: our first participant is choosing her own words and using neural activity to communicate them in real time,” said Matt Angle, PhD, Paradromics’ founder and CEO. “She determines what she wants to say, and the system enables her to say it.”

Recommended Reading

Aug. 13, 2026 News by Marisa Horak, MS

AI voice analysis may help doctors spot early signs of ALS, study suggests

System uses 421 microelectrodes, each smaller than a human hair

Motor neuron diseases include ALS, a neurodegenerative condition marked by dysfunction and loss of motor neurons, the nerve cells responsible for controlling muscle movement. Over time, people with ALS may lose the ability to move, speak, swallow, and breathe.

Implanted just beneath the brain’s surface, Connexus BCI is designed to record electrical activity from individual nerve cells in the brain’s motor cortex. According to Connexus, the system uses 421 microelectrodes, each smaller than a human hair, to capture neural signals.

The recorded neural signals travel through a flexible connector to a chest transceiver. The information is then transmitted wirelessly to an external computer.

Artificial intelligence and neural-decoding algorithms analyze the recorded patterns to determine what the person intends to say. The system can then translate those signals into synthesized speech, text, or computer commands.

Recommended Reading

Aug. 7, 2026 News by Marisa Horak, MS

Study finds traumatic brain injury may be sign, not cause, of ALS

‘Okay, I have a lot to say’

The first participant, a Michigan woman in her 60s, had severe dysarthria, a motor speech disorder caused by damage to the nervous system or weakness in the muscles that control speech, and retained limited speech. Although her husband could often understand her, communicating with other people had become much more problematic.

During a task that allowed her to communicate freely, she generated the sentence, “Okay, I have a lot to say.” Connexus decoded the sentence in real time, with no errors, according to Paradromics. She also used the system to communicate with their grandchildren during a live phone call, asking questions and responding through speech generated by the device.

Later, Matthew Willsey, MD, PhD, a Connect-One study investigator and neurosurgeon, asked what she would say to a participant considering a BCI and the same surgery she went through. “Don’t be afraid,” she responded through Connexus. “It’s a wonderful experience to be able to help others by your participation in the research study.”

“The participant and the Paradromics team have achieved a major milestone by demonstrating the ability to speak freely through a fully implantable device designed for clinical use. Restoring communication may soon be available to others with diseases impacting the ability to speak,” Willsey said.

Data from this first participant showed distinct patterns of brain activity associated with attempted speech, or trying to produce speech, and imagined speech, which is when a person generates language internally without physically attempting to speak.

“Across populations of individual neurons, we can see structure associated with listening, preparing to speak, attempted speech and imagined speech,” said Vikash Gilja, PhD, Paradromics’ chief scientific officer. He added that longer-term recordings will help researchers understand how stable these signals are and how the participant and decoding system adapt to one another.

In addition to synthesized speech and text, Connexus is designed to allow users to control computers and other digital devices. The FDA recently approved an expansion of the study to include connections to personal computing devices such as phones, tablets, and laptops.