Frequency-Lowering and Audible Bandwidth: Interview with Ryan W. McCreery, PhD
Beck: Great to catch up with you.
McCreery: Thanks, Doug. Nice to speak with you, too.
Beck: Ryan, if you don't mind, I'd like to get your thoughts as to where we are as a profession, with regard to fitting frequency-lowering technologies. It seems a good place to start that discussion is on the bandwidth research from a few years back?
McCreery: Sure. Well as you know, for females and small children, listeners need audibility often up to 9000 Hz to differentiate /s/ versus /sh/. And so the question became how do you accomplish that audibility?
Beck: Right. And if I recall correctly, the back story was that Pat Stelmachowitz and her team wrote about this in Ear and Hearing, perhaps in 2007? She showed that for hearing-impaired children using a 5000 Hz bandwidth, fewer than a third of them could accurately identify the /s/ sound. However, if you used a 10,000 Hz bandwidth, almost 90 percent of the children with hearing loss could accurately identify the same /s/ sound. Further, the American Academy of Audiology recommends that for newborns and children, a minimum bandwidth of 9000 Hz is the goal.
McCreery: Yes, that's right. And the thing to remember is that bandwidth will be limited by a number of factors, not the least of which is the patient's degree of hearing loss. For example, if you have a great hearing aid with a bandwidth going to 9 or 10K for speech, but the patient has a severe hearing loss, it's not likely you'll be able to deliver 9000 Hz to that patient. In fact, there was a paper we published in JAAA recently that made these same points (Kimlinger et al, 2015). The variability of the ANSI estimate of bandwidth is an issue with speech stimuli, and so it's reasonable to evaluate bandwidth as a function of hearing loss, and that's how we ended up looking at frequency lowering as an option.
Beck: And so how do you evaluate frequency lowering with regard to these issues?
McCreery: Frequency lowering has to be evaluated through the lens of audibility. That is, optimizing frequency lowering, means to use the smallest amount of frequency lowering possible, to obtain the largest improvement in audibility.
Beck: And to be clear, audibility means how much of the speech signal is available to the person wearing hearing aids. But to take this a bit further is your idea of using the "audible bandwidth." Can you explain that?
McCreery: One of the challenges when using the Speech Intelligibility Index (SII, or the other published audibility or articulation indexes) is to quantify high frequency audibility is that the bands above 4000 Hz have little, or very limited importance in the calculation. And so you can have very large differences in audibility in the high frequencies, with little change in the SII.
Beck: So the tools we have are limited and in-exact. No news there! But what can we do to get around the limited measurement tools?
McCreery: That's where the audible bandwidth comes in. We wanted to account for differences in the high frequencies, which is clinically useful and something that can be easily used. So with audible bandwidth, we're looking at where the average speech spectrum and the peak of the speech spectrum cross the audiogram, from the verification screen.
Beck: And so you're looking at the highest frequency that can be heard through a particular hearing aid, programmed a certain way?
McCreery: Exactly. In fact, the reference listed below will help explain the audible bandwidth, using the Verifit. And so audible bandwidth helps us understand what the patient can hear without frequency lowering, and what they should hear with frequency lowering. So if the person can perceive sound out to 8000 Hz, it's just not likely that frequency lowering is going to provide additional information or benefit. Further, and this is a point you make in your lectures, Doug, when we apply frequency lowering, we're distorting the natural speech signal. And so we have to be careful about the balance between adding audibility and increasing distortion.
Beck: And it seems that in addition to distorting spectral cues, we're potentially distorting interaural timing and interaural loudness differences.
McCreery: Perhaps, but these are some of the issues clinicians should consider when applying frequency lowering strategies. And further, let's be clear, if you're applying frequency lowering appropriately and conservatively, you're not moving heaven and earth. You're improving a little audibility some of the time, that is, you're potentially capturing an additional 10 percent of speech audibility, some of the time. But again, we have to be careful when we say ten percent, as ten percent in the very high frequencies may be much more valuable than ten percent from other spectral ranges.
Beck: Right, we cannot presume equal weighting. Further, any programming changes made via the manufacturer's software are suspect as they may not reflect what is actually going on in the patient's ear canal, and the only way to actually know is to measure it through real ear measures.
McCreery: Yes, and as we all know, most people fitting hearing aids don't bother with verification or validation or real ear measures, and so these potentially amazing technologies are very likely fit inaccurately.
Beck: And to be clear, I think it's true that neither you nor I are opposed to frequency lowering. There may indeed be times when it is useful, but without verification and validation and without real-ear measures, the likelihood of a wrong fitting program is greater than the likelihood of a correct fitting.
McCreery: I believe that's a fair point. Let's face it, if you haven't accounted for your patient's actual ear canal resonance, you're guessing, and the opportunity to guess correctly is very slight.
Beck: I agree. Ryan, it's been a pleasure chatting with you and I encourage our colleagues to take a good look at your audible bandwidth template and to use it when considering audibility versus distortion.
McCreery: Thanks, Doug. I appreciate your interest in our work.
Beck: My pleasure, Ryan. Thanks for your time.
Ryan W. McCreery, PhD, is the director of the Center for Audiology at Boys Town National Research Hospital, Omaha, Nebraska.
Douglas L. Beck, AuD, Board Certified in Audiology, is the Web content editor for the American Academy of Audiology and the director of public relations with Oticon, Inc.
For More Information, References, and Recommendations
Kimlinger C, McCreery R, Lewis D. (2015) High-Frequency Audibility: The Effects of Audiometric Configuration, Stimulus Type, and Device. Journal of the American Academy of Audiology 26(2):128–137.