With the aid of Artificial Intelligence Technology, a team of scientists at the University of California, San Francisco, has recently developed a brain implant that is capable of reading people’s mind and translating their thoughts into speech. The discovery is ‘life-changing’ as it could be helping those who suffer from neurological impairment diseases or those who have lost their ability to speak.

In context, deciphering speech from neural activity is a challenge as it requires a precise understanding of the rapid synchronization and control of our vocal tract articulators like our mouth, tongue etc. For this to be possible, a technical device must encrypt sound representations in human cortical activity into representations of articulatory movement which is then transformed in speech acoustics.

Incredibly, similar studies like this have been attempted by other researchers. For example, a team from Columbia University performed a study that revealed neurological patterns produced in the brain when we speak, listen or even imagine speaking or listening. They used a digital device, a vocoder, to recognize and synthesize speech based on activity recorded in the brain.

What researchers in the University of California have achieved is decoding what is interpreted as speech within the brain and translating it into actual words. This was done by using an implant, which is an electrode reading the electrical activity in the brain.  Instead of translating the pattern depicted as code for each word spoken like other researchers have done with limited success, this team tried a slightly different approach. The implant that they used recorded the signals in the brain that manoeuvre the voice box, the jaw etc. Then, it transcribed this data as movement made from different sounds.

One of the researchers, Professor Edward Chang, commented;

“For the first time, this study demonstrates that we can generate entire spoken sentences based on an individual’s brain activity.” … “This is an exhilarating proof of principle that, with technology that is already within reach, we should be able to build a device that is clinically viable in patients with speech loss. »

However, as the implant is still a working prototype, further testing must be done to tweak discrepancies in functionality. For example, because the implant relies on brain data prescribed by articulatory movements, only patients with certain diseases can be treated: motor neurone disease, brain injuries, throat cancer, Parkinson’s, multiple scleroses, cerebral palsy… Regrettably, at this point in time, patients with certain types of strokes who cannot communicate will be unable to benefit from this technological find.

Despite some limitations, the implant study is still praised by experts as a positive starter for a forward way of thinking in the healthcare industry. University College London Professor, Sophie Scott, acclaimed;

« This is very interesting work from a great lab, but it must be noted that it is at very early stages and is not close to clinical applications yet. »

References:

1. ‘Exhilarating’ implant turns thoughts to speech, April 2019, BBC News Health
2. Speech synthesis from neutral decoding of spoken sentences, April 2019, Nature
3. Towards reconstructing intelligible speech from the human auditory cortex, Jan 2019, Nature 
4. AI technology could turn thoughts into speech, February 2019, ScienceFocus