Researchers at the University of California, Davis, have developed an innovative brain-computer interface (BCI) capable of translating brain signals in real time from people who have lost their speech as a result of neurodegenerative diseases.
The breakthrough has been tested on a patient with amyotrophic lateral sclerosis (ALS), enabling him not only to communicate verbally using a synthesized voice, but also to modulate intonation and even sing.
The development of this new BCI has been published in the journal Nature and represents a milestone in neurotechnology by combining brain implants and artificial intelligence algorithms to restore a natural, instantaneous and human form of communication.
The researchers explain in the article that for the first time they have succeeded in enabling a man suffering from amyotrophic lateral sclerosis (ALS), who had lost his ability to speak, to regain the ability to communicate orally.
This was achieved through a brain-voice neuroprosthesis that instantly synthesizes the man’s voice; the BCI also allowed him to modulate his synthesized voice in real time, change intonation and sing short melodies.
This new interface is capable of translating and reproducing the patient’s brain signals through a speaker at a speed similar to that of human speech.
Until now, what brain implants did was to transcribe on a screen what patients who had lost the ability to speak were trying to express. This procedure involves a delay between the detection of the brain signal and its transcription onto the computer screen, which makes communication difficult.
This BCI consists of four microelectrodes implanted in the brain’s temporal region, which is responsible for speech. After recording neuronal activity, it is transferred to a computer that interprets the signals, generating language that will be reproduced through speakers, all at high speed.
In addition, the patient was also able to modulate his intonation to ask questions or emphasize certain words in a sentence, and he even sang short songs.
Thanks to the BCI, listeners were able to understand almost 60% of the words spoken with the synthesized voice, which was a major advance given that before the BCI, the patient could understand only 4% of the words.
Neuroscience and Artificial Intelligence
Advanced artificial intelligence algorithms were used to instantly translate brain activity into synthesized speech.
The algorithms were trained with data collected while the participant attempted to speak sentences written on a computer screen, which provided the researchers with information related to what he was trying to say.
This brain activity profile showed the firing patterns of hundreds of neurons, which the researchers aligned with the speech sounds the participant was attempting to produce. This helped the algorithm learn to accurately reconstruct the patient’s voice from his neural signals.
As we have already published in the Bioethics Observatory, several brain implants have previously been used, successfully establishing communication patterns that allow brain activity to be interpreted and transferred to external devices.
The interface now unveiled represents a breakthrough because, when applied to ALS patients, it would enable them to cope with the increasing difficulty in communicating that occurs over the course of the disease. This factor could significantly improve their quality of life.

Julio Tudela and Ester Bosch
Bioethics Observatory – Institute of Life Sciences
Universidad Católica de Valencia