This scan speech of the brain reflects activity in the cortex – a part of the frontal lobe involved in speech production.Credit: Montreal Neurological Institute/Science Photo Library
A brain-decorating transplant that translates nerve signals into audio speech, has allowed a woman to hear with paralysis that she intends to say almost immediately.
Researchers extended the device – known as a brain -computer interface (BCI) – with Artificial Intelligence (AI) algorithms, who used to think of the woman about them, and then speak them loudly using a synthetic voice. Unlike previous attempts, users can only produce sounds only after finishing the entire sentence, the current approach can detect words together and can turn them into speech within three seconds.
Brain-reading devices allow paralyzed people to talk using their ideas
Conclusions, published in Nature Neurology On 31 March1The BCI represents a large step towards which is of practical use.
The old speech-generating BCIs are similar to “a WhatsApp conversation”, Christian Herf, a computational neurocystist at Mastrich University, Netherlands, who were not involved in work. “I write a sentence, you write a sentence and you need some time to write the sentence again … It does not flow just like a normal conversation.”
The BCIS that streams the speech in real time, research has the “next level” because they allow users to give voices and emphasis that is characterized by natural speech.
Brain-component reader
Study participants, N, lost their ability to speak in 2005 after a stroke. A few 18 years later, he performed a surgery to place a paper-split rectangle containing 253 electrodes on the surface on his brain cortex. Transplantation can record the joint activity of thousands of neurons at the same time.
Researchers individually voiced the synthetic voice like N’s own voice before their injury, by training the AI algorithm on recording from their wedding videos.
