People with dentures may soon be able to speak normally, thanks to researchers who have developed dentures fitted with sensors that record pressure exerted by the tongue and may uncover the tongue's hidden movements in speech production.
This may help the researchers to design better voice synthesisers and also false teeth and braces that interfere less with speech.
"The aim is to try to understand how humans are able to speak by modelling the speech-production apparatus," New Scientist quoted Yohan Payan, a researcher at the TIMC lab near Grenoble, France, and part of the team who worked on this project.
He said that it was very difficult for them to know how much pressure is exerted by the tongue on the teeth while producing some speech, for example when making a "T" sound.
"This closure of the vocal tract allows you to pronounce this consonant. To model this, you have to be able to estimate the level of force applied by the tongue," he explained.
Earlier, such measurement required the researchers to stick sensors to people's teeth, or embedding them into an artificial palate, which actually hampered the normal functioning of the tongue.
But, to avert this problem, the French team opted to hide their sensors inside dentures made for 20 volunteers who had already lost their teeth.
They specifically designed individual devices for each patient, having one or two sensors embedded inside. Then they were arranged on the palette for recording tongue pressure while pronouncing particular consonants. The output was transmitted to a computer via a wire running along the inside of the cheek, but away from the tongue. On the other hand, the sounds made by a person were simultaneously recorded using a microphone.
As they did not change the physiology of the mouth, it enabled the volunteers to speak normally at the time of the measurements. They recited tongue twisters for generating the results.
"This is a neat trick; a new twist on a methodology that has been around for some time. The idea of using denture patients in this way is clever," said Joe Perkell, a researcher in MIT's Speech Communication Group in Cambridge, Massachusetts, US.
Till date, the researchers have published results on the production of the sound "T", and now they are planning to do the same for other vocalisations.
In fact, Payan is hopeful that ultimately they will synthesise more realistic human speech than is currently possible.
"When you hear a [voice synthesiser], you can recognise that it's not a human voice, it's a kind of humanoid voice," he said.
Also, with a better understanding of the mechanisms of speech, they will be able to integrate the unique features of human speech into these models. Payan said that the work may also help in designing dentures and orthodontic braces that have less impact on a patient's ability to speak normally.