Scientists in the US are working on a device that can decode brain activity to spell out sentences. 

‘Neuroprostheses’ and brain-computer interfaces (BCIs) are devices that replace lost nervous system function, and have the potential to restore communication to people who cannot speak or type due to paralysis. 

However, it has been unclear whether silent attempts to speak can be used to control a communication neuroprosthesis. 

Previously, neuroprosthetic systems have been able to decode a handful of words from brain activity, but they were limited to a specific vocabulary and the participant had to attempt to speak the words out loud, which required significant effort.

A team from the University of California has now designed a neuroprosthesis capable of translating brain activity into single letters to spell out full sentences in real time, and demonstrated its use in a participant who suffered from limited communication because of severe vocal and limb paralysis. 

In tests, the device was able to decode the brain activity of the participant as they attempted to silently speak each letter phonetically to produce sentences from a 1,152-word vocabulary at a speed of 29.4 characters per minute, and an average character error rate of 6.13 per cent. 

In further experiments, the authors found that the approach generalised to large vocabularies containing over 9,000 words, averaging a 8.23 per cent error rate.

They were able to demonstrate real-time decoding of silent attempts to say 26 alphabetic code words from the NATO phonetic alphabet in a highly highly accurate and rapid way. 

Brian activity was obtained from an implanted 128-channel electrocorticography (ECoG) array and used to train classification and detection models.

The results highlight the potential of silently controlled speech neuroprostheses to generate sentences through a spelling-based approach using phonetic code words. 

The researchers say further work is required to demonstrate if this approach is reproducible in more participants.

The latest study is accessible here.