University Place

cc >> Hi, everyone. I'm a researcher here at the Waisman Center and a professor in the Department of Communicative Disorders, and our lab is the Learning to Talk lab. I'm going to talk about sound and word learning in children with cochlear implants. You're going to hear some of the same things that you heard in some of the other talks. Like Ruth I'm going to start with the glass half empty or the glass half full, I'm going to call it good news versus bad news. Cochlear implants are one of the trials of modern engineering and single processing. Children with cochlear implants have much better speech and language skills than we ever could have imagined looking at children a generation ago with the same type of hearing impairment who wore hearing aids. They are mainstreamed, they speak beautifully, their language is excellent. We never would have imagined that we would come so far so fast. We also have to acknowledge that we're not there yet. Children with cochlear implants as a group, we know there's a lot of individual variability, but as a group they don't perform as well as their same age peers with normal hearing. Speech is slightly less intelligible. On every measure of academic achievement, they don't do quite as well. Here I'm talking also to know that a lot of our kids with cochlear implants live, like in Lake Woebegone, all children are above average. We're not content if 100 is the average score, we're not content that our children with cochlear implants have scores at the average. We want them to be above average like all the other children in their school or town. Of course here we have to think of what comparison group we're going to use. We don't want to compare apples to oranges. It depends who we compare them to. If we compare children with cochlear implants, if we compare their speech or language, if we compare them to children of the same chronological age, they're not going to do as well. It makes sense to compare them to these kids, they're in the same classrooms, but it's not fair to compare them to these kids. For one thing, they've been hearing for less time. Even if you had a cochlear implant at age 1, when you go to kindergarten at age 5, you've been hearing for four years, where everyone else has been hearing for five years. We could instead compare them to children of the same hearing age. We could take kids and say, okay, you're five, but you've only been hearing for four years, so we're going to see how you do compared to kids of 4 years. When we do that, they often do better. That makes sense. Even if you get your cochlear implant at age 1, you don't start learning everything at age 1. Maybe you start learning to speak and learn words, but you've been learning lots of things from when you were born to age 1. The difference between the concepts of an apple and orange, even if you don't know the words apple and orange. What we often do in our lab is compare them to children with the same size vocabulary. And when we compare them to children who have the same vocabulary size we find the most similarities between children with cochlear implants and children with normal hearing. The problem is that children with cochlear implants tend to be on the average of six months delayed in vocabulary. So they're not quite there. Now, you know, as people have already said, it's obvious why we have these differences. You know, the auditory input isn't as good. 22 channels versus thousands. So they don't have as good auditory input, and they also haven't been hearing for as long. As people have been saying, we have many, many advances in these areas that are going on even as we speak. So what we're doing in our -- you know, we're not engineers, we're not surgeons in my lab. So what we're doing is focusing on two things. We're trying to understand how speech and language development is different for children with cochlear implants as we compare them to children with normal hearing. This is both in terms of producing sounds and perceiving sounds. Also in terms of learning words. And we're also interested in individual differences in performance. So if you remember back to Ruth's slides, you can see there was a huge range of performance in children with cochlear implants. Some children are really what we call stars. They perform as well or better than children with normal hearing. And what we'd like to do is think of -- try to understand as well as we can what is going on with this -- with this variability, so that we can potentially make all children into stars, all children performing as well as children with normal hearing. So before I talk about a few of the things we found, I want to show you this diagram which talks about the interactions between perceiving speech, producing speech, and learning words. So let's start here with speech perception. This is going to be things like being able to hear the difference between a ka and a ta, or a and ay, and this is where we expect children with cochlear implants to have problems. The auditory implant isn't as good. And hearing the difference between ka and ta is really hard. The problem here is that this interacts with everything else. So speech perception interacts with speech production. Being able to make the difference between a ka and ta or an ah and ay. The better your speech production is, the better your speech perception will be, and the better your speech perception is, the better your speech production will be. We have these interactions, and the more, the better you get at perception and production, the better memories you're going to have with these individual speech sounds. Now if we look up here at words, you know, having memories for words, you really have two kinds of memories. You have word meanings. Obviously we know the difference between something you keep a bird in and something you eat on a birthday, but when we have a memory for a word, we don't just have the memory of its meaning, we also have the memory of its sounds. We also know that cake is different than cage because one ends in k, and one ends in ja. The better memories you have no speech sounds, the better you'll do at learning and remembering words. We could have this diagram go even farther up. Up here would be things like reading and writing, things important for academic achievement. Because we know that the two I think that's most important for being a good reader are knowing lots of words, and being able to break those words down into sounds. You see this one weak link in speech perception can lead to problems, difficulties all the way up. This is what we're interested in. I can show you a couple results that we have that show similarities and differences between children with cochlear implants and children with normal hearing. So this is from work that a graduate student did, and what she was interested in was how children produce sounds. She took the two sounds, sa and cha, like sip and chip, and she recorded them from children with cochlear implants, and children with normal hearing of the same chronological age. She was measuring the acoustics in the sounds, and one of the big differences has to do with the frequencies between the sounds of sa and cha. Sa has a higher frequency than cha. You see the results here for children with cochlear implants and children with normal hearing. This measure of frequency is along the vertical axis. Sa frequency results are shown in the white bars, frequencies for cha are shown in the gray bars. There were all productions that listeners heard, and said that sounds like a good sound. You can see for both groups, there's a higher frequency for sa than for cha, but you can see that the two sounds are much more distinguished in the children with normal hearing than the children with cochlear implants. There's a more robust contrast. There's a bigger contrast between the two sounds in the children with normal hearing. Maybe it's things like this, maybe it's these subtle differences that lead to the speech of children with cochlear implants being just a little less intelligible. The other thing that is interesting here, the -- these black lines are the average, but we've also plotted data for every individual child in these circles. You see, first of all, there's lots of variability for both groups. These are 2 and 3-year-olds. You can also see there's at least one child with a cochlear implant who is performing just as well as the best children with normal hearing. That shows you there are these stars. That's where we want everyone to be. In this next slide we're comparing children with cochlear implants to children who have the same size vocabulary. And this experiment had to do with learning novel words. If I said a word you to, this is actually a word, but it might be an novel word to you, pulchritude. The first thing you'd have to do is encode it enough that you could repeat it. Whether you know this word or not, you would be good as repeating it. It's like attitude, latitude. It's not hard to remember and say pulchritude. This is something we've known no a long time, the larger your vocabulary the better you can repeat a new word, and that is something that is very important for word learning. We did a similar experiment, not with pulchritude, but with made up words, and we got exactly that result. We got the receptive vocabulary, how many words they understand along this axis, how accurate they were in repeating novel words along this axis. You see for both groups as their vocabulary increases, they're more accurate. What is really cool about this slide, we plotted the data for children with cochlear implants in black. The data for children with normal hearing in red, and you see the exact same relationship. Once they get that vocabulary size they're just as good as repeating novel words. That's a very exciting result. The a little bit of concern here is that when we match children for vocabulary size to children with cochlear implants were on the average about six months younger. When they get to that vocabulary size they're functioning, they're learning new words at least to the point of repeating them in just the same way that children with normal hearing are. So that's exciting. This result is for adults. And this is from work of Matt Wynn, who is in the audience. His picture is here so you can find him later. He's a post-doc, who's working with Ruth and me. This is work he did with other colleagues at the University of Maryland when he was a graduate student. He was interested in the same sounds, sa and cha, in perception of those sounds, and telling them apart both in adults with normal hearing and adults with cochlear implants. He gave them two conditions. The listeners heard both male and female voices, and he gave them two conditions. He gave them one condition where they saw a female face with a female voice, and a male face with a male voice, and in the other condition they just heard the voices, and didn't see the faces. He found that the adults with cochlear implants were more sensitive to the visual cues. They paid more attention to the faces than the adults with normal hearing. In a way this makes sense. If the auditory signal isn't as good you should be trying to use other cues that you might not have to have if you have normal hearing. So what they were doing, they were paying attention to a cue that might be in a normal hearing listener redundant, or was an important cue for a listener with cochlear implants because the auditory input is degraded. So with these results, I say the million dollars question is how can these results inform clinical progress. I wouldn't say we have a lot of answers yet, but one thing I think we have to do, particularly looking at Matt's results is thinking about how we do auditory verbal therapy. In auditory verbal therapy the focus is I'm getting the child to pay attention to just what they're hearing to the extent that we don't even provide them with facial cues. There certainly are advantages to that in that we really want kids to concentrate on the auditory input that they haven't had up to when their cochlear implant is activated. But it could be that we're carrying it a bit too far in not also teaching children to integrate visual cues. Maybe we need to have more than one condition for children in auditory verbal therapy where sometimes they're getting the auditory cues only, and sometimes getting the auditory and the visual cues. Because we know, we suspect that to be a successful cochlear implant user, you're going to have to use all the cues that you have at your disposal, because the auditory input is degraded. I want to switch gears a little bit now and talk about another approach that might help all of our children be stars. And here I want to talk about something that is a little further away from speech and word learning, and that is executive function. Executive function is what psychologists call the cognitive processes involved in the regulation of three things. One of these things is working memory. Working memory is a little different than short-term memory. Working memory has both stored space and processing space. And working memory capacity is limited, as I'm sure we've all experienced. The example I like to give is how you can drive and talk at the same time until the driving gets bad. All of a sudden you're in a lot of traffic or it starts pouring down rain, and you say to the person sitting next you to, I can't talk anymore, I have to concentrate on driving. Cognitive flexibility is being able to switch from one task to another. Any of you in here with children know that is something that develops gradually. That's why we give our children warning about transitions. Five more minutes, and it's going to be time to get ready for bed. Inhibitory control is exactly what it sounds like. It's learning to inhibit our impulses. Instead of grabbing the last cookie on the plate, look around at everyone else at the table saying, does anyone else want that cookie? We know all these things inhibitory control, flexibility, and working memory, develop gradually from childhood to adulthood. What's really interesting about this is that executive function seems to be involved in speech perception and in word learning. So in our lab we've seen that adults who have better executive function, and we can measure this by standardized tasks, pay more attention to redundant cues and speech perception. If you take again that example of sa versus cha, the primary cue in English, the difference between cha and sa is difference in frequencies. There's other cues. They're not as important in English, but they're there. Another important cue is the frequency at the beginning of the vowel. What we find is that people who have poor executive function as measured by tasks tend to pay attention only to that primary cue, whereas people with better executive function pay attention to both the primary cue and the secondary cues. Other people, David Pisoni at Indiana University has shown that children and adults who have better executive function measured by standardized tasks or sometimes by questionnaires are better able to recognize spoken words. If you ask someone to listen to a passage, and press a button every time they hear the word cup, people with better executive function who have cochlear implants will be faster and more accurate at that task. So this suggests that possibly teaching -- you know, having people, training people to have better executive function might result in better outcomes with cochlear implants in terms of children and possibly adults. Now what is exciting about that, if you recall, think back to the original talk that Teresa Palumbo gave, the Center for Improving Health and Minds, which is right here at the Waisman Center has developed a preschool program to improve executive function. It's similar, I don't know if anyone hear has done as an adult mindfulness stress reduction. The program is very similar to a mindfulness based stress reduction program, except it's age appropriate for preschoolers. It starts with children lying on their back, and they watch a rock rise and fall on their stomach as they're taking belly breaths then they're asked to hold up a finger when they hear a bell stop ringing, and so on. It's a six week program for about, I don't know, half an hour a day, and it's been shown to be effective in improving executive function in preschoolers and early school age kids. It'd be really exciting to do such a program with children with cochlear implants, and see if that does translate into better speech perception or word learning for those kids. So I'd like to thank my wonderful collaborators for their generous support we've had for this research, and I can take questions.

applause