The device created by a group of scientists at the University of California, San Francisco, USA, will translate brain activity into speech and will help people who lost speech from neurological disorders. Gopala Anumanchipalli, a former dual degree doctoral student in Language Technologies at Instituto Superior Técnico (IST)/ Carnegie Mellon University (CMU) and member of this group of scientists helped us to better understand this promising research.
“Many neurological disorders, such as Amyotrophic Lateral Sclerosis (ALS), cerebral vascular accidents or other types of palsy, often lack control of their speech muscles. The current auxiliary devices, based on eye tracking and muscle spasms, are slow and result in very low communication rates, such as 10 words per minute at best”, explains the former Técnico student. Nowhere near natural speech, which rushes by at around 150 words per minute. And for that reason this research group thought about the possibility of helping these patients to get their voice back.
“We recorded high-density electrocorticography (ECoG) signals from five participants who underwent intracranial monitoring for epilepsy treatment as they spoke several hundreds of sentences aloud. Our findings represent one step forward for addressing a major challenge posed by patients who are paralysed and cannot speak”, explains the Técnico alumnus. The study was published in Nature research journal last April.
The team trained one participant who did not vocalize, but instead just silently mouthed sentences. “We were thrilled to see that we can produce speech even in this case. We expect that individuals with speech impairments will regain the ability to speak, as long as their brains’ speech centers remain intact”, says the Técnico alumnus.
“We began to work in this project 5 years ago. Improving the life quality of these patients was our main motivation”, says Gopala Anumanchipalli. “This is the starting point for the creation of a brain-machine interface that will allow individuals with neurological disorders to regain their ability to speak. This research is just the beginning (…) but its potential applications can be huge”, explains the researcher at the University of California.
According to the Técnico alumnus, the skills acquired during his PhD were vital “to understand the speech signal” and allowed him “to gain excellent engineering skills that were useful to carry out this research work”. “My PhD research consisted of
prosody modeling techniques to improve computer-generated speech, and I had a deep interest in understanding the physiological basis of human speech production”, he adds.
