St. Olaf News
Students work to create a training program for cochlear implant users
July 9, 2014
Why is it that some recipients of cochlear implants are able to understand speech very well, while others struggle to make sense of even the simplest sentences?
This is the question a new project through St. Olaf College’s Collaborative Undergraduate Research and Inquiry (CURI) program is investigating this summer.
“Although most implanted individuals do acquire the ability to perceive sound, there is substantial variability in how well cochlear implant users can understand spoken language,” says St. Olaf Assistant Professor of Psychology Jeremy Loebach, who is leading the research.
“This variability cannot be accounted for by differences in the cause of their hearing loss, the onset and duration of deafness, how old they were when they received their implant, or other surgical or physiological factors,” he adds. “Despite extensive research efforts, understanding why this variability exists and how to control it remains a significant clinical problem.”
Loebach is working with Rachel Bash ’15 and Brandon Cash ’16 to discover the cause of this variability. But more than that, they want to develop a way to train cochlear implant users how to improve their hearing.
“We believe that explicit training will provide implantees with a foundational set of neurocognitive skills that they can use to better develop auditory and speech processing proficiency, thereby reducing some of the variability observed in outcome and benefit,” Loebach says. “Essentially, we want to teach them how to hear.”
In search of a solution
Cochlear implants are electronic neuroprosthetic devices that are surgically implanted in the inner ear and interface with an externally worn microphone. By converting sound into electrical impulses and transmitting them directly to the auditory nerve, they enable people with severe hearing loss or profound deafness to artificially perceive sound.
This high-tech treatment might bring to mind the visor that restored sight to the fictional character Geordi La Forge in the television series Star Trek: The Next Generation. But unlike La Forge’s visor, which in the sci-fi series enabled him to see far beyond the spectrum of light visible to the naked eye, cochlear implants do not increase the range of sound users can perceive.
In fact, the electrode array that carries the signal has a maximum of 22 channels to represent sounds with frequencies between 20 and 20,000 Hz. A normal hearing ear, on the other hand, has some 3,500 hair cells to cover the same range. This means that cochlear implant users have to do “a lot more with a lot less information,” according to Loebach.
But Loebach hypothesizes that in addition to this reduced range, the variability seen in the success of cochlear implant users may be caused by the lack of any standardized training paradigm, particularly for adults.
“With children there’s a lot of rehabilitation,” Loebach says, “and they will often have teams of audiologists following them throughout their early developmental years. But most adults who get a cochlear implant after they’ve lost their hearing get nothing. They’re left on their own to figure out how to learn how to hear again.”
In order to address this problem, Loebach and his students have been testing cochlear implant users in order to illuminate what it is that those who are successful are doing differently from those who are struggling. Armed with that information, the group can then modify their training program to specifically target those differences.
“The approach that we’re taking is that if we can better understand what the people who are doing really well are doing, and whether that’s something about the implant, the training, or about just how they understand the world and perceive sound, then we can make a standardized training program to help the people who are struggling to do better,” Loebach says.
Testing and training
The test is a computer program that works by playing recordings of different auditory stimuli, from ambient noises like a cash register closing to individual words and sentences. The cochlear implant user repeats back what they think they heard, and is then shown the correct response while the sample plays again. One important set of stimuli are anomalous sentences — sentences that are grammatically correct but have no meaning, like Noam Chomsky’s famous example “Colorless green ideas sleep furiously.”
Loebach includes these anomalous sentences because they force listeners to rely on their sense of hearing to identify the words, not on linguistic context.
“When people have severe hearing loss, they tend to rely on context more and more,” he says. “We find that many cochlear implant users do indeed use this strategy, and perform very well on meaningful sentences. But for the anomalous sentences, context cannot help you to identify the words. Many cochlear implant users struggle with such sentences, and including such stimuli in training will teach the user to rely on their hearing, rather than the sentence context. So by removing the linguistic context of the stimuli, we can see what acoustic information cochlear implant users can make use of, and how this compares to their speech perception abilities.”
Loebach and his students hope to test at least 100 cochlear implant users over the course of the year. With that data, they will be able to develop their training model and put it into action.
Above is a recording of an anomalous sentence — “Plans now are a stiff pin for landing” — that has been altered to approximate what it might sound like to cochlear implant users. Below is the unaltered recording.
A multidisciplinary collaboration
In addition to his project with Bash and Cash, Loebach is also working with Katie Berg ’15 on a separate project through a Magnus the Good Fellowship. He and Berg are investigating whether musical training can benefit speech production and perception in cochlear implant users.
The study is looking at both pediatric and adult cochlear implant users, and will use several exercises to work on improving pitch discrimination abilities, an area where most cochlear implant users struggle. Loebach and Berg hope that improvement in pitch discrimination will also help cochlear implant users improve their speech intelligibility and production skills.
“I like that these projects combine music and psychology,” Berg says. “It’s the perfect mix of psychology and music. I didn’t ever think I would use my aural skills knowledge in the field of psychology, let alone to help someone else, so that’s pretty exciting.”
Cash says he was fascinated by the idea of teaching people how to hear.
“I’ve always loved music, and so I’ve always thought that hearing is important, because that’s how I enjoy what I love,” he says. “I took the Psychology of Hearing course with Professor Loebach, and then he brought up cochlear implants and the idea of helping people to hear again, to relearn how to hear. It just seemed like such an appealing idea to me. It’s related enough to my studies that I’m interested, and it’s far enough away that I feel like I’m expanding and growing.”
What interested Bash about the project was that it combined research with working with people.
“I wanted to get a sense of what research was like, as well as what it’s like to work with people,” she says. “I’m trying to decide whether I want to go into research versus something more with people, and I think this will help me decide. And this is starting to become a real passion of mine too, so I might even consider doing something with speech and language in the future.”