Magnets give a voice to the larynx-less

A few small magnets and a headset that interprets facial movements could one day help people without voiceboxes speak again.

For people who have lost their voicebox, a device that interprets facial movements could help them speak again, New Scientist reports.

Currently, many people who’ve lost their larynx (to cancer, injury or otherwise) are often fitted with a valve in their throat that diverts air from their lungs to the esophagus when they exhale. But these valves can become clogged shortly after use.

So this new device (pictured) recognizes whenever the wearer mouths a word. "We can pick up information about the way they are moving their lips, teeth and tongue around, and from that information reconstruct their speech," says researcher Phil Green of the University of Sheffield.

It uses small magnets placed into the mouth and on the tongue to create a magnetic field. Sensors in an external headset detect changes in the magnetic field as the person mouths the words.

So far, the system can recognize only about 50 words. And the team plans to develop magnets that can be implanted into the tongue, according to team leader James Gilbert at the University of Hull. The researchers are also aiming to reduce the size of the headset down to something like a Bluetooth device.

People who suffer a stroke or those with cerebral palsy, Parkinson's disease, or motor neuron disease might not lose their speech entirely, but they can lose the ability to speak coherently.

Just as human listeners have trouble understanding impaired speech, so do conventional speech-recognition systems, says Stuart Cunningham also at the University of Sheffield. His team is developing an iPad-sized prototype that could learn to recognize impaired speech and replay a clearer version.

After the system recognizes the word, it’s replayed by a voice synthesizer.

Alternatively, a voice recording can be taken from the patients themselves before their impairment becomes too severe, and then used to adapt a standard artificial voice.

The researchers plan to begin testing on people in the next 12 months.

Via New Scientist.

Image: James Gilbert, University of Hull

Related on SmartPlanet:

This post was originally published on