Email
Password
Remember meForgot password?
    Log in with Twitter

article imageScientists extract speech directly from the brain

By Tim Sandle     Apr 26, 2019 in Science
San Fransisco - Scientists have extracted speech directly from the brain, as part of a technique designed to one day help people with sever medical conditions to speak. The process is a speech synthesis technology, designed to recreate speech patterns.
The research to date has been with health participants, using a process that taps directly into the brain. The technology has several iterations to go through before it can be used on patients. However, the results of the neuroscience experiment to date show sufficient promise, for further development to be undertaken.
Technology is designed to translate neural activity into speech. This approach would be transformative for people who cannot communicate due to a neurological impairments. To achieve this, technology is required to decode speech from neural activity. This is based on precise and rapid multi-dimensional control of vocal tract articulators.
The researchers, from University of California San Francisco, have designed a prototype neural decoder which can create kinematic and sound representations, encoded in human cortical activity, in order to create synthesized audible speech. The biggest challenge with the neuroprosthetic technology is recording cortical activity and transforming this into interpretable representations of speech acoustics.
The trials undertaken on healthy subjects used an array of intracranial electrodes to record brain activity (that is, electrodes implanted into the brain). These were used to help the researchers to better understand the complex relationship between the movements of the vocal tract and the speech sounds.
The following video explains about the technology in further detail:
Data from the healthy subjects, who were required to read hundreds of words, enabled the scientists to understand the vocal tract movements required to produce the particular sounds required for the formation of different words, including pressing lips together in a certain way or tightening vocal cords in a given fashion, or shifting the tip of the tongue in a particular direction. This data will provide the basis of the next phase of the development.
Lead researcher Professor Edward Chang says about his attempts to restore fluent speech to patients: "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."
The research has been published in the journal Nature, with the paper titled "Speech synthesis from neural decoding of spoken sentences."
More about Speech, speech synethesis, Brain
More news from
Latest News
Top News