They announced advances in brain implants that restore speech to people

Researchers specializing in neurosurgery from the universities of Stanford and California, in the United States, announced relevant advances in two projects based on the introduction of brain implants to restore speech to people who have lost it, according to an article published this Wednesday in the specialized magazine Nature.

First case

The first case described in the scientific document is that of Pat Bennet, a 68-year-old woman who did office work, maintained a healthy and sporty lifestyle, but ten years ago was diagnosed with Charcot's disease, a neurodegenerative disorder that deprives progressively to the patient of his movements, until causing him the complete paralysis.

As a consequence of this pathology, the woman began to have difficulty expressing herself and over time she completely lost the ability to speak. In this case, the specialists introduced four small chips with 64 microelectrodes, made of silicone, in March 2022.

These devices, embedded in 1.5 millimeters of the cerebral cortex, record the electrical signals produced by the parts of the brain dedicated to language, messages that are carried out of the brain through a set of cables and analyzed with an algorithm, which in the last four months "learned" to interpret its meaning.

In this way, the algorithm associates the signals with phonemes and thus composes words, with the help of a language model and a device through which Bennet speaks through a screen at a rate of more than 60 words per minute, according to what he reported. the AFP news agency.

"Now we can imagine that in the future it will be possible to restore a fluent conversation in a person suffering from paralysis" of language, Frank Willett, a professor at Stanford and co-author of the study, said at a press conference.

This represents an ability to speak less than 150 or 200 words per minute for a normal person, but it is a rate three times higher than that achieved by a person with language paralysis in 2021, in another scientific experience promoted by the same laboratory. .

second case

The second case in the Nature article was carried out by Edward Chang's team at the University of California and consists of a device based on a band of electrodes installed in the cerebral cortex.

Its efficiency is comparable to that of the Stanford study, with an average of 78 words per minute, and it represented a huge advance for the patient in whom the electrodes were installed.

Having suffered paraplegia due to a brain haemorrhage, until now his speaking capacity was 14 words per minute, with the help of a technique that followed head movements.

One of the specificities of the Californian laboratory's device is that it not only analyzes the electrical signals from the part of the brain dedicated to language, but also the movements of the lips, tongue and other facial muscles that produce sounds.

"Between five and six years ago we began to understand the electrical networks that govern the movements of the lips, jaw and tongue and that allow us to produce the specific sounds of each consonant, vowel and word," explains Professor Chang.

The interface between the brain and the device in this laboratory not only produces language in the form of text on a screen, but also with an artificial voice and an avatar that reproduces the patient's facial expressions when speaking.

Consequently, the laboratory at this California university is now aiming to develop a wireless version of the same mechanism. This would have "profound consequences for the independence and social interactions" of the patient, points out David Moses, co-author of the study and professor of neurosurgery at the University of San Francisco.