Researchers at the California Institute of Technology demonstrated at the 2022 Society for Neuroscience conference in San Diego that they could use a brain-machine interface (BMI) to predict the words a quadriplegic patient was thinking.

BMI devices are brain implants that could help patients who have lost the ability to speak. What this interface does is translate brain signals into words or audio output. It therefore makes communication possible for people who have lost their speech abilities due to diseases or injuries.

Advances in the decoding of vocalized, attempted, and imitated speech have been very significant. However, results in decoding internal speech are scarce and much work still needs to be done to achieve high functionality.

Therefore, this scientists’ proposal is one of the first of its kind and can be found in an article published in the journal MedRxiv, along with the study’s findings.

How Does the Brain-Machine Interface Work?

BMI technology was first used to help people with quadriplegia control robotic arms and hands. But in the current study, the same technology was used in the area of language and communication to reconstruct speech.

In previous studies, it was possible to predict participants’ speech by analyzing brain signals recorded in motor areas when a participant whispered or imitated words.

But this recent research went further and managed to predict what someone was thinking — that is, their internal dialogue. The accuracy they achieved was one of the highest on record.

To achieve that level of prediction, the researchers recorded brain signals from individual neurons in an area of the brain called the supramarginal gyrus, located in the posterior parietal cortex.

First, there was a 15-minute training period in which the researchers tested the BMI device to recognize the brain patterns produced when the quadriplegic participant spoke or internally thought certain words. They then displayed a word on a screen and asked the participant to say it internally. The BMI algorithms managed to predict eight words with an accuracy of up to 91%.

Although the work is still preliminary, the researchers hope that this technology could help reconstruct speech in patients with brain injuries, paralysis, or diseases such as amyotrophic lateral sclerosis (ALS).

Reference:

Online internal speech decoding from single neurons in a human participant


Discover more from TekCrispy

Subscribe to get the latest posts sent to your email.