Join the club for FREE to access the whole archive and other member benefits.

Translation of brain signals from a paralyzed patient into text by machine learning

15-Jul-2021

Key points from article :

People who can’t speak or write because of paralysis have limited ways of communicating.

Edward Chang, who led the work says “... we’re at the very beginning of a new chapter, a new field”.

Researchers developed a “speech neuroprosthetic” for decoding the brainwaves.

Which controls the vocal tract, muscle movements and larynx that form each consonant and vowel.

The man who volunteered to test the device was in his late 30s.

Computer analyzed the patterns when he attempted to say common words.

Prompted to questions like “How are you today?” to the answer “I am very good”.

They started him to try to say specific sentences rather than answering open-ended questions.

David Moses, lead author says "It takes about three to four seconds for the word to appear on the screen..."

Next steps include improving the device’s speed, accuracy and vocabulary size.

Research by University of California, San Francisco published in The New England Journal of Medicine.

Neuroprosthesis provides natural communication for people who cannot talk

Mentioned in this article:

Tap on icon for description, click on resource name for more details.

Academic

Neurosurgeon at UC San Francisco.

Journal

Scientific Journal devoted to medical research.

Academia

Public research university that is part of the University of California system and dedicated entirely to health science