Sound Science: Li Xu examines how cochlear implants can be fine-tuned for speakers of tonal languages
By Adam Liebendorfer
For the tens of millions of Mandarin Chinese speakers who suffer from hearing impairment, the tonal language can pose serious comprehension challenges. The Mandarin word “ma” inflected one way means “mother,” but, with a slight change in the speaker’s tone, “ma” can mean “horse.”
Although more Chinese are using cochlear implants, electronic devices surgically attached to the inner ear to enhance sound, the current technology can’t capture the subtle rise and fall of tonal languages such as Mandarin.
Ohio University researcher Li Xu has spent the last decade exploring how to improve the devices, not only for speakers of China’s official tongue, but for the millions of tonal language users around the world.
According to the U.S. Food and Drug Administration, as of December 2010, approximately 219,000 people worldwide have received cochlear implants. China is a rapid growing market for the devices, due to the large population and the widespread use of antibiotics that are known to cause hearing impairment in children, Xu explains.
Cochlear implants work by transmitting a signal through electrodes to the cochlea, a spiral-shaped part of the inner ear. There the electrical signal stimulates auditory nerves for hearing. Current models of cochlear implants have devices attached outside of the head that house a microphone, a processor, and a transmitter. A receiver and electrodes are surgically implanted inside the skull and connect wirelessly to the transmitter. Xu says future models will be entirely internal, which will allow patients to participate in activities such as swimming underwater.
Cochlear implants, however, cannot convey the subtle pitch changes found in tonal languages. The cochlea has hundreds of hair cells and hundreds of different channels of sound to transmit to the brain. Understanding English only requires eight channels, but Xu has determined that comprehending tonal languages requires 30. The best implants available have 22 electrodes and can cover 22 channels, but even they do little to help cochlear implant users. Nerve cells tend to die off when they fall into disuse, as is the case with many patients with hearing impairment. Xu says this makes active cells “few and far between.”
“And the problem is, even when you have 22 electrodes, the best patients can only use up to 8 channels because of this,” he says.
Difficulties with recognizing pitch also impact a patient’s ability to listen to music, Xu adds. Music requires hundreds of channels, and current implants can’t capture music’s tonal richness.
Much of Xu’s most recent research, funded by the National Institutes of Health, seeks to overcome these obstacles. Each summer, he travels to several cochlear implant clinics in China to determine how children with the devices can perceive and reproduce tones. The first test he conducts asks children to pair pictures with sounds they hear. To study how they can produce pitches, Xu records the children speaking or singing in their native Mandarin and then analyzes the recordings’ acoustics.
Xu found that many children with cochlear implants can typically get around their inability to perceive tones by using context in the conversations. However, since they do not pick up the pitch differences, they do not speak with them. Other native speakers have difficulty understanding these children. In contrast with the tonal roller coaster of Mandarin, cochlear implant patients sound monotone.
Nevertheless, Xu says the quality of life of the children with implants “has changed tremendously.” Hearing impairment can be viewed as a burden to one’s family more often than it would be in the United States, he says, and few Chinese know sign language. With cochlear implants, children are no longer sequestered in schools for the deaf, he adds, and they now have a chance to go to college.
At $15,000 to 20,000 an implant, however, it’s been difficult for Chinese citizens who aren’t wealthy urbanites to afford the device. The Chinese government has started allocating public funding for low-income Chinese, and Taiwan recently donated enough support for 15,000 patients to get implants. These efforts are usually confined to China’s biggest cities, though. Xu pegs the number of potential Chinese cochlear implant users in the millions.
Xu’s colleague Phillip Loizou of the University of Texas-Dallas, who also is exploring technologies to overcome tonal recognition problems, notes that Mandarin speakers don’t feel they can justify the price of an implant that still can’t relay pitch.
“We are still not very sure how to get those extra channels to the brain,” Xu admits.
Researchers are exploring several possible solutions: Growing more neurons in the ear that can hear a wider range of pitches, changing how the implant connects to the brain, and cleaning up the sound that implant users hear.
Unlike individuals with normal hearing, implant users hear a wall of sound and must figure out how to distinguish someone’s voice from the background noise. To counteract this, researchers are looking into a combination of hardware and software changes to the exterior device that transmits the signal to the brain. One approach is to include adjustable settings, such as one for in the car or one for at a restaurant, that help cancel out the background noise, Loizou says.
While Xu contemplates these technological challenges to tackle in the next phase of his research, he’s active in the international community, organizing last year’s Asia-Pacific Symposium on Cochlear Implants. China will host 2015’s symposium, which Xu thinks will do one of the most needed things for tonal language speakers with hearing impairment: raise awareness.
“Education has caught up a little bit, too,” he says. “A lot of people don’t know implants can help.”
This article appears in the Spring/Summer 2012 issue of Ohio University's Perspectives magazine.