Purdue researchers have developed an in-ear device that uses recorded chatter to improve the speech of Parkinson's patients
Speech pathologists have been trying to get their patients with Parkinson’s disease to raise their voices for years. Although the condition is primarily characterized by tremors and difficulty in walking, most patients also suffer from speech problems, particularly slurring and what’s known in the field as weak voice. Now, Jessica Huber, an associate professor in the speech, language, and hearing sciences department at Purdue University, in West Lafayette, Ind., has developed a wearable device specifically designed to get Parkinson’s patients to speak louder and clearer.
Speech pathologists usually work with Parkinson’s patients one-on-one, but the patients often regress when they leave the clinic. Huber wanted to get patients to talk louder in both clinical and natural settings. She knew that audiologists often used recorded party chatter to test patients’ abilities to raise their voices in response to noise, a reflex action known as the Lombard effect. Huber reasoned that if she could integrate the party chatter recording into a portable and user-friendly package, she would be able to activate the Lombard reflex in her Parkinson’s patients anytime, anywhere.
Huber doubted it was possible to engineer such a device at the university level, however. ”It was kind of a joke at first,” she remembers. Then, in 2007, she talked to researchers in Purdue’s biomedical engineering department who told her that building the device would not only be possible but also relatively easy and inexpensive. ”The device itself is not extremely sophisticated,” says James Jones, the lead engineer for the project. ”The novel part is the application.”
Huber and Jones started collaborating in late 2008, and by February 2009, they had a product to test on patients. The device consists of a small accelerometer that sits on the throat and is wired to an earpiece about the size of a hearing aid. The accelerometer senses when the patient is speaking and signals, via analog circuitry, the earpiece to play the party chatter (the recording is a product named Multi-talker Babble, from the St. Louis–based company Auditec). Jones and Huber made sure that the chatter was loud enough to trigger the Lombard reflex but not so loud that it prevented patients from hearing. The device is active only when the patient is talking, which keeps the interruption at a minimum and helps save battery life.
Although the project is still in the early stages—Jones and his team have built about 30 devices so far—Huber is encouraged by the results. The idea of having party noise playing in your ear every time you speak might sound annoying, but Huber and Jones say the overall response has been overwhelmingly positive. One subject even lovingly refers to the voices as ”her people,” says Jones. And after a few months, some patients were able to speak louder without the device, says Huber. ”It’s worked for all the study participants to some degree,” she says. ”The challenge now is to figure out why the device works for some people better than others.”
To answer this question, Huber has been analyzing usage data from the device, together with physiological data from external sensors placed around patients’ ribs and mouths. She expects to have a better understanding of exactly how the technology—and by extension, the Lombard reflex—works by next year.
In the meantime, Huber hopes her research will help increase awareness about the speech deficits associated with Parkinson’s disease. ”Problems with speech volume and articulation may not be the predominant symptoms [of Parkinson’s], but they’re still a major issue both for the patients and their caregivers,” she says.
Listen to what Parkinson's patients hear with the experimental Purdue University device.
About the Author
Erica Westly is a freelance science writer based in Brooklyn, N.Y. Her article ”Sign Language by Cellphone” appeared in IEEE Spectrum Online in March 2009.