Home > News > Speech Device Could Assist Palsy Victims

Speech Device Could Assist Palsy Victims

From Daily Progress
July 6, 1990

By Lane Thomasson

A flick of the wrist or nod of the head will one day allow people with cerebral palsy and other disorders to speak through a computerized speech synthesis device, University of Virginia researchers hope.

The device under development at UVa is design to electronically reproduce sounds of the human voice and allow speech-impaired people to speak with fluidity and inflection.

"This will be a major breakthrough and could benefit people around the world," said Randy Pausch, an assistant professor of computer science at the UVa.

"Last year we developed the speech synthesizer and this year we have devised a novel interfacing technique which will allow the patients to use the sound. Over the last 12 months we've gone from 'Can we do it' to 'Yes, we can-here's how.'"

"Children with cerebral palsy are all completely different," Pausch said. "They're supposed to adapt to a world they cannot control. But we're making the world adapt to them. We can take whatever movement they can give us and we make it easy because the device is part of them."

The speech synthesizer, now known as the Tailor project, operates on a person's individual range of movement, which a programmer or therapist tracks into a computer program. Once the movement, such as swinging the arm or wiggling the foot, is recorded, it is overlaid onto a sound grid. The vowel and consonant sounds are produced as the user moves a tracking device through the grid.

The synthesizer can reproduce any Romance language, Pausch said.

Stroke victims and patients with Parkinson's disease and multiple sclerosis are potential users of the device.

"These people have live minds but uncooperative bodies, and they are unable to communicate because they often cannot control their vocal cords or tongues and have too little movement to type or write," Pausch said. "previously, all that was available to them was a spelling board...and that only let them convey words a letter at a time or a maximum of 20 symbols."

The speech synthesizer devised at UVa is the first one capable of reproducing inflection and pitch change, so it sounds human, Pausch said.

"No one wants to sound like Darth Vader, even if that is the only way to speak. This project will let people sound human rather than mechanical," he said.

The synthesizer is like a musical instrument, Pausch said - anyone can play, but it takes years to master. First attempts to speak using the synthesizer could be frustrating, as a patient would have to get used to the sound grid and process of physically combining sounds through body motions. "But they will have to learn speech this way just as a baby does - it's a developmental process," Pausch said.

The exact cost of the synthesizer has not been determined, he said.

"We're talking hundreds of dollars - not thousands," Pausch said.

The device is scheduled to be ready for use by 1998, Pausch said.

Therapists and others at the Kluge Children's Rehabilitation Center are helping Pausch and his graduate students with the research. Children with cerebral palsy and other defects are testing the device in the engineering school lab.

"We are trying to help these people who have a need for communication, not just mobility," Pausch said. "It's fine if you can manipulate your wheelchair - but where does that get you if you can't talk to anyone once you get where you set out to go? Short of healing these people, this is the single best thing you can do for them - give them the ability to speak."

Original Article | Local Copy


More news about Randy Pausch


Return To List