Nov 2020 ABC News
Cochlear implant users can usually understand voices well in quiet conditions, but find noisy environments harder
Cochlear implants are so widespread that it's easy to take for granted what a life-changing tool they are for people with hearing loss. But they're not perfect. Implants don't always perform well in noisy areas, and they can be more effective for some people than others — but improving the design is a labour-intensive process that relies heavily on the work of volunteers who use them.
Now, a biomedical engineer at The University of Sydney, who has cochlear implants himself, has come up with an algorithm that provides a more streamlined way of testing new ideas so it is easier to tailor them to an individual's ability to hear different noises.
Greg Watkins, a former electrical engineer, first noticed his hearing going about 15 years ago. He got hearing aids, but his hearing continued to decline until he was profoundly deaf. "If I stood next to a jet engine, the world was a silent place," Dr Watkins says. His hearing loss inspired him to go back to university and study biomedical engineering, and he ended up doing a PhD into cochlear implants. His studies gave him the impetus to put theory into practice, and get the implants himself.
A cochlear implant involves wiring going into the spiral-shaped cochlea in the inner ear
While hearing aids more or less amplify the sounds around you, cochlear implants work by directly stimulating the nerve in the inner ear with electrical impulses in a way that mimics the stimulation that we'd usually get from sounds. Dr Watkins had day surgery to have the implant put in, then four weeks later was back at the audiologist's office for "switch-on day”. "Eventually they turn on the microphones and you hear the world around you," he says. "But it's not the world you used to know. Sounds are quite different and you need to learn to hear again. But to hear it all was amazing."
Greg Watkins prepares for his cochlear implant to be switched on for the first time
Cochlear implants are powerful technology, but how well those electrical impulses translate into sounds varies a lot between individuals, says Amanda Fullerton. a Sydney-based audiologist who's doing her PhD on how the brain changes after cochlear implantation. "We often say, we hear with our ears and we listen with our brain," Ms Fullerton says. "The signal still has to leave the ear to get to the brain and the brain has to integrate that signal and interpret it. There could be various sources of variability between people from the level of the ear right up to the brain. In a quiet situation many cochlear implant users can hear quite well, but it's when there are multiple people talking or there's background noise, that there can be a problem."
So while cochlear implants can be hugely useful, there's still room for improvement. Researchers are coming up with new ideas for improving the implants all the time, but testing them is time consuming and labour intensive. Volunteers with cochlear implants have to spend hours in quiet rooms listening to recorded sentences and repeating them back. Then researchers score these results and use them to draw conclusions about what is and isn't working. Dr Watkins has been one of these volunteers. "That approach works, but it takes time. It's really tiring. I've done it myself. And as a recipient, it is really hard work," he says.
Greg Watkins has combined his first-hand experience as a cochlear implant user with his research
Dr Watkins's engineer mind clicked into gear. Surely, he thought, there had to be a way to make this process better? He and his collaborators at the University of Sydney and Cochlear have come up with an algorithm to help predict the results of these tests for different people. Their software tool can be used by researchers to narrow down which approaches are worth pursuing, potentially reducing the number of hours human volunteers are needed for. "If we were able to predict how well someone would hear with a particular sound processing approach, then we'd be able to look at how well new ideas might work for that person," Dr Watkins says. "If we take the test results for a cochlear implant recipient in one [noise] condition, we're able to predict how well they will be able to understand speech in a range of other conditions."
While Dr Watkins's tool is designed for researchers, it will lead to a broader range of options when it comes to configuring a cochlear implant for a recipient, says Ms Fullerton. "We know that different people have different abilities to hear in noise, and that isn't always predictable, so if there was an algorithm that could be adjusted to an individual's ability to hearing noise, then that could presumably assist them in in hearing better in a range of environments," she says.