07/18/2021 at 10:00 CEST
Researchers at Oregon Health & Science University have developed a neural implant that can read a person’s thoughts and convert them into written words.
It is specially designed for people who for whatever reason have lost the ability to communicate: just try to speak out loud for the mechanism to work.
The system captures the intention of the person, decodes their thoughts into specific words and reflects them on a screen that another person can read.
The technology is still not perfect, as the algorithm that performs this feat succeeds in about 50 percent of the time.
However, the level of communication increases when the person thinks of phrases that we repeat frequently and that have been previously written on a screen.
In addition, the system has a function similar to that of auto-correction that increases the precision of both individual words and complete sentences.
The technology has been tested in a human patient who had been paralyzed for 20 years and represents an innovation: it is the first time that language has been deciphered from the electrical signals that are generated in the speech area of the brain.
The system translates signals intended to control the muscles of the vocal system to speak words, rather than signals to move the arm or hand to allow the writing of a sentence, the researchers said in a statement. This approach takes advantage of the natural and fluid aspects of speech and promises faster and more organic communication, they say.
Related topic: It is now possible to write thoughts only with the mind
Human historyThe story of this experience is recounted by The New York Times: it says that three years ago, this patient, now 38 years old, agreed to work with researchers in neuroscience.
His life had changed in 2003 due to a car accident that required surgical intervention. After undergoing surgery, he suffered a stroke and was left in a coma and unable to speak.
For years, he used other technologies to communicate with enormous difficulty, until scientists implanted a rectangular plate of 128 electrodes, designed to detect signals from sensory and motor processes related to speech and the mouth, lips, jaw, the tongue and larynx.
Through 50 sessions developed over more than a year and a half, everything the patient wanted to say was transferred to a computer via a cable connected to his head.
The system translates brain activity that would normally have controlled your vocal tract directly into words and sentences. You can communicate at 15 to 18 words per minute, although researchers think the technology can be fine-tuned to the pace of a typical conversation, which is 150 words per minute.
The researchers also emphasize that the system improves with experience: at first it confuses words and expresses some thoughts erroneously, but over time it improves the translation of words because the patient’s brain gets used to expressing itself in this way.
Does not read minds
Does not read mindsThe researchers clarify that the electrodes do not read the mind of the patient, but rather detect the brain signals corresponding to each word he tries to say.
The system does not capture random thoughts, but the thoughts associated with each word.
The researchers emphasize that this technology is aimed at helping people who lack the ability to speak and who preserve neural pathways related to oral language.
This technology would therefore be of great help to people with brain injuries or conditions such as amyotrophic lateral sclerosis (ALS) or cerebral palsy, who have insufficient muscle control to speak, notes NYT.
In an article published in The New England Journal of Medicine, the architects of this technological development write: “Technology to restore the ability to communicate in paralyzed people who cannot speak has the potential to improve autonomy and quality of life. An approach that decodes words and sentences directly from the cerebral cortical activity of these patients may represent an advance over existing methods for assisted communication. “
ReferenceNeuroprosthesis for Decoding Speech in a Paralyzed Person with Anarthria. David A. Moses et al. N Engl J Med 2021; 385: 217-227, July 15, 2021. DOI: 10.1056 / NEJMoa2027540
Top photo: Gerd Altmann on pixabay.