 So, before I start, what I'd like to do is acknowledge basically all the people who did the actual work on this project that I'm representing here in this talk. So we've had a lot of activity, a lot of people doing great work here at NAL, and they are really the ones responsible for what I'm about to show you today. So the project that I'm going to talk about really derived from the approach that NAL takes with everything that we do, which is we start by trying to understand the unmet needs of people with hearing loss, whether that means children in classrooms, cochlear implant wearers, people with unilateral hearing loss. We approach everything we do from an unmet needs in order to understand what solutions we should be providing for them, and that's what drove this project. So what was the unmet need that we saw that kicked off our work? Well, let me change this to full screen here, and let me get out of your way. So the main unmet need that drove us was our observation in the middle of last year, 2020, with everyone wearing masks that people with hearing loss were struggling more than most because of this, as everyone knows people with hearing loss, including people with hearing aids, have a more difficult time understanding speech than people with normal hearing, and this situation seemed to be making things worse. So you see a scattering of headlines here from a variety of news articles highlighting this problem. So we didn't dive in right away to solve the problem, but we wanted to understand it a little more and make sure that this was an extensive problem that we should be spending time on. So what was the situation that was driving this unmet need? Well, first of all, face masks were being mandated in many situations, public transportation, social gatherings, medical environments. We were finding, even in Australia, where COVID was not as prevalent as elsewhere, that mask wearing was being required in a lot of situations where communication was important. Secondly, a lot of these situations were actually critical to understand the details and the nuances of what was being said, particularly in healthcare settings where you don't want to get a recommendation from the clinician wrong. You want to be really careful to make sure that the patient understands everything that's being said. And again, people with hearing loss already are struggling in these situations now when the healthcare providers wearing masks became even more of a concern. Even when masks weren't being mandated, governments were encouraging people to still wear masks when they could. And so even everyday communication with friends in social situations became challenging. And then you also had situations where because of those communication difficulties, people were choosing not to wear masks. So in order to be heard more clearly, they were removing the mask, which was putting people at risk in an unsafe situation now, increasing the likelihood of the spread of the COVID virus. So these were the sort of our initial observations of the need. And then finally, even for people wearing masks, if there was the introduction of some masks with transparent mouths so you can see the lit movement, but they weren't widely used and acoustically they weren't very good. So what is the impact of wearing masks on speech? Well, I mean, we we intuded without any evidence that we were pretty sure that the masks were going to attenuate parts of speech, particularly the high frequencies. We also determined that this would impact speech understanding in noisy environments. So even though the mask, the mask is reducing the speech to noise ratio by the talker, so that along the way, it will make it worse. But in noisy environments with reverberation, it's the high frequencies where you have a better speech to noise ratio. And these are the frequencies that were getting hurt the worst. We also recognize that lip reading is an important part of how people understand others in noisy environments, particularly people with hearing loss wearing the mask loses those cues. And we knew that was going to have an impact on people's understanding. And then we determined that speech production was actually not that impacted by wearing masks based on a literature review. But we weren't too concerned that the production of speech was being impacted by masks that were being worn. So it was really about the acoustic effects and not on the impact on how people were able to articulate the speech that they were saying. So who are the people experiencing this problem? Well, first of all, everyone was having problems understanding people wearing masks. But again, this is particularly problematic for people with hearing loss and people wearing both unaided and aided because they already struggled to understand speech and they already have difficulty in the high frequencies where the masks were having the biggest problem. And the possibility that the masks were reducing speech cues to below their levels of ability was a concern, even wearing hearing aids. We also saw from talking to people that because of social distancing, often someone with a hearing loss will bring a partner with them, especially in a critical communication situation such as at a hospital, a healthcare environment. And because of social distancing, often the partner wasn't even allowed into that facility. So this made it even more difficult for people who had hearing difficulties to begin with that someone who they rely on quite a bit to make sure that things are heard and understand we're not there for them. And you can see some of the quotes that we have here. So and I'll just read a couple here for you. Now I only go grocery shopping with someone with me. I knew I was going to be hearing was going to be hard, but this is a whole new level. So at least when grocery shopping, you can bring a partner. And you know, with face masks, I had issues over the phone, over the counter, I needed to talk louder. And the one at the bottom attending a clinic appointment this week, which I struggle to understand what was being said to me. So all of these were our challenges that people were experiencing day by day. So what was the evidence that we had that this was really a problem besides our intuition and and sort of empirical some, some circumstantial evidence. Well, if we look at the publications, you know, there was a there was a several publications that had been released that identified hard numbers around these problems. So here's one from the American Journal of Otolaryngology, showing that over 60% of people experience moderate to severe difficulty understanding speech by those in emergency rooms when they had mild to profound hearing loss. Again, transparent masks weren't very good. This is a publication about the National Health Service in England, where a significant number of them experienced communication difficulties, wearing the surgical mask, and reported if difficulties even wearing a shield, which I think many people thought would be better than a mask, turn out not to be the case. Our own measures of the speech articulation speech intelligibility index or the articulation index indicated that mask wearing was going to be a problem. We know that lip reading can help people by by effectively improving the speech to noise ratio by four to six dB. And therefore removing them is almost like hurting or reducing the speech to noise ratio by that same amount. In here in Australia, the Department of Health, just as a matter of policy will allow hearing aids to be fit on people who have normal audiograms if they have visual impairment because of the acknowledgement that lip reading is so important to understand speech. So if you have very mild hearing loss along with vision loss, it's acknowledged that you are going to struggle more an indication that loss of lip reading really is an impairment in communication. And then finally, we noticed that badges being sold in England that say hello, I'm deaf. Masks make it hard for me to communicate. They were sold out. So there were people with hearing loss, who were struggling who are wearing badges buying badges and wearing them to inform other people of their own difficulty. So we saw a lot of evidence out there that this was an unmet need that really demanded our attention. We also looked at Reddit, where we found a thread where someone asked people with hearing impairment how corona mask policies affected their daily lives. So you see a word cloud here of what we saw. But here's some here's some direct quotes of what we saw on the Reddit thread. So I won't read them all out to you. But as you sort of look around and scan these quotes here, it's very clear that it was affecting people's lives in many ways in their stress, in their emotion, anxiety, being afraid to go out because of the difficulty that they were having. And so once again, the experience of hearing difficulty is not just reduced to acoustics and audibility of speech cues, but impacts their whole lives, their whole psychosocial situation. And again, this really emphasized to us the need to address this. So we reached out to some of those Reddit people to ask them some very specific questions about the challenges that they were having. We got 107 respondents, 41% of which were hearing aid and CI wares. So when we asked them up here, you know, what top middle, when you were talking to someone who was wearing masks, do you have trouble understanding? 79% reported difficulty, 86% if they were hearing aid wares. When you're talking with someone who wasn't wearing masks, you have trouble understanding well, a much smaller number, right? So this demonstrates that the mask really is the cause of the difficulty that people are having. The main concern that most people had was lip reading, but many people had concerns about the muffled sound they're experiencing as well. People were frustrated and 62% of respondents reported avoiding interacting with people with masks. So again, this is a situation that really shouldn't be tolerated. And what strategies were people using? You know, people were asking people wearing masks to talk louder, asking them to repeat themselves, asking them to write things down. People were coming up with their own solutions as people do when they're faced with unmet needs that aren't being solved for them. So that's sort of the full sort of list of evidence that we accumulated before we decided to develop our solution. So what we thought we would do is do some acoustic measurements to understand what the impact of masks on speech understanding is, develop a recommendation for audiologists on how they can help their clients who are hearing aid wearers hear better when wearing masks. So create a mask preset or a mask program in their hearing aids that will compensate for the acoustic effects of masks and hopefully make their experience better, overcome those anxieties, social isolation that was resulting from this situation and then validate our recommendations and make sure they're having the effect, positive effect that we thought they would. So first we started by measuring the acoustic effect of masks. You can see our setup here. We had a hats system, which is a telecommunication mannequin used very extensively in the mobile phone and, you know, earpiece, headphone world and so on. Then we had Kemar, our friend and mannequin sitting in the middle of that room. And then we measured the impact of various types of masks with various kind of situations, both with hearing aids and without hearing aids. And so just I want to show you at a high level what we saw because this is I think in a this difference between masks and face shields has a lot of impact and what I'm going to talk about next. So here you see the amount of attenuation provided by wearing a mask or a face shield. The mask is that flatter line in blue. The face shield is this peaked one here in orange. And what you're seeing here plotted is the change in sound level DBSPL between our mannequin not wearing a mask and mannequin wearing a mask as measured in the ear of Kemar. So I think everyone is familiar with these types of curves as well. But what struck us immediately was how face shields actually boosted speech in these mid frequencies by up to 6 dB and then how much more significantly they attended attenuated speech in the higher frequencies. So that's really important to keep in mind. Some of the frequencies were actually boosted but the ones that were attenuated were significantly more attenuated than with masks. So we measured a variety of different masks that were in that people were using out in the real world that we could get our hands on. So here you see four masks for different curves. You know the surgical mask had the least impact shown here. Some of these other masks had more significant significant impact. And you can see the average there in green. But they all sort of show the same thing. Just a couple DB or so below a thousand hertz and above a thousand hertz it the attenuation increases to a maximum of around four kilohertz anywhere between you know four and nine decibels of attenuation at the peak. So what did we do. Well it's this is a this is a linear effect on the levels of speech received. So it seemed by applying some linear amplification we could compensate for this. And so what what did we do. We calculated the attenuation at the automatic frequencies from several of these masks. So which one did we choose. We thought well we could come up with recommendations for different masks but the hearing aid wear is going to be exposed to different people wearing different kind of masks. So it didn't make sense to have a mask specific recommendation. And we didn't want to take the average because the average would depend on the mask that we happen to have selected in that group that we measured. So what we did is we calculated the range of attenuation at each of these automatic frequencies and then we selected the midpoint and said sir this is you know not the median but the midpoint of attenuation at that frequency. And we used that to determine the amount of gain that would be necessary to compensate for a wide variety of different masks. And we didn't develop end up developing a recommendation for face shields. And this is in part because of that very large attenuation that occurred at the higher frequencies. And I'll get to that a little bit more near the end of the presentation. So these are the recommend gain recommendations that we came up with from one to six dB of additional gain in a mass setting at these frequencies. And you know we didn't try to limit these to make them smaller but I was glad that they didn't turn out to be too large because this gain was not only going to be applied to people wearing masks but to all other sounds in the environment. So you don't want to over amplify sounds all the other noise in the environment which may make it uncomfortable for the hearing aid to wear. But we put hearing aids on chemar and we measured the effect of mass aided on chemar. Then we created a program that had this gain boost in the hearing aids and we showed that the audibility of speech was restored back to normal with these gain settings. So something that should be obvious but we did the measurements anyways just to confirm that our recommendations restored the sound levels to the proper levels of audibility. So why didn't we make recommendations for face shields or the combination of face shields and masks which some people were wearing. Well first of all as you can see in that figure again it would have required a recommendation of up to 20 dB of additional gain in the hearing aid and to me that didn't make any sense at all. So sure you'd be restoring speech of those frequencies to normal but you'd now be adding 20 dB of gain to everything else in the environment and that just seemed like a disaster waiting to happen. We just knew that that was a non-starter. No one was going to accept that. Secondly in order to restore audibility to normal we are actually going to have to prescribe that gain be reduced in the mid frequencies and recommending that you reduce frequencies is never a good thing for I think someone wearing a hearing aid. They're already struggling with audibility and for those who have open fittings you're not going to be able to control the frequencies in these lower frequencies anyways. So no matter what we prescribe you're not going to be able to turn them down. So because they're going to come through in the unamplified level. So you really just couldn't control it with gain. And then finally as you'll see we actually didn't measure any impact of on speech understanding from face shield wearing. So we didn't see a need and sort of this unmet need and you had the lip cues so we didn't see an unmet need that people had for compensation for people wearing a face shield. So we stuck just to masks. So in terms of validating our recommendations we recruited 12 subjects here is their audiograms sort of a range of hearing losses and we what we did is we they were hearing aid wearers and we had them wear hearing aids with and without the boost in order to measure the change in a variety of measures. So we had them listen to the hats wearing them wearing a mask and also without a mask measure their speech understanding and measured subjective quality measures of with those subjects. So here's here's what we found. So first of all the amount of speech degradation in terms of the speech reception threshold. So we measured the speech understanding the SNR at 75% correct. This figure shows you the impact of either face masks on the left face masks and face shields in the middle or face shields only on the right on speech understanding compared to no face masks and no face shields. So zero means that the face mask or face shield had no impact on speech understanding. Here you see a two to three dB worsening of the speech reception threshold for these subjects or of the SNR. And here with the face mask and face shield you see a similar impact on speech understanding but you see quite a spread from subject to subject some up to five dB impact which be quite significant on ability to understand speech. Here we see on average there was no impact on speech understanding from face shields and statistically if we look at these three different conditions face mask and face mask plus face shield were statistically different from not wearing a mask or a shield but this was not face shield was not. So another reason why we ended up ignoring face shields with our recommendations again based on because some of the face shields actually boosted audibility in in addition to to reducing it so it seemed to sort of be a wash in terms of the same path. So the amount of benefit when we compensated for when we added our recommendations for face mask or for the face shields we actually implemented a perfect compensation for face masks and face shields here. We saw that the face mask condition got significantly better in speech understanding. The face mask plus face shield did not when we added the high frequency boost and the face shield alone speech understanding actually got worse when we added that high frequency boost and we're not sure why it got worse it could be that we were hitting the limits of the hearing aid or some other internal sort of rollover effect of speech understanding at high levels. So again another reason to stick to face masks only our gain compensation provided benefit only in that situation. We had people we asked them about the acceptability of the loudness of sounds with our boost we were worried that people might not tolerate the loudness that they heard. So this is the loudness acceptability of the target speech with the boost in different conditions. So first of all in the no masking condition with no with no boost you got loudness you know around zero to minus one they were happy with it presumably the very bottom is strongly disagree that the loudness is acceptable the very top they strongly agree that loudness is acceptable and here you see when you at when the face mask condition with the boost the tolerability the acceptability of loudness has improved so that they agree that it's acceptable face mask and face shield also would improve with the face shield there was no real change to acceptability of the loudness of target speech. Now when we look at the loudness of environmental sounds we see that again with with no mask loudness of environmental sounds is fine with face mask and our boost we see that the loudness was also fine although quite a spread in from subject to subject face mask with face shield or face shield we see that the loudness got worse that the acceptability of the loudness of environmental sounds got worse which is what we expected because that boost was being applied to everyday sounds and making them louder than they should be for that person. We also in addition to the gain recommendations provided recommendations for other settings for the hearing. So first of all once the gain adjustments are made the audiologist needs to try to hit those targets as best they can knowing that every fitting software is different every hearing it is different some fitting software won't let you meet those targets exactly but you want to get as close as you can maybe even use some real ear measurements to make sure that you're hitting that those additional gain top of the target. You also want to make sure of course once you add this added gain that you have it introduce feedback as a problem. You know these these are all sort of standard audiology practice when you are making adjustments to a hearing. The audiologist should also ensure that when they add that gain that environmental sounds aren't going to be too loud for their client. You could do that by you know banging some things in your clinic and so on to see if they can be tolerable with this added gain. You also want to make sure that you're not hitting any of the limits of the hearing aids saying maximum power output or saturation or you know some other limitation the hearing aid may have at higher levels of gain or higher output levels that they may hit with this additional gain. So again pretty standard practice when fitting hearing aids. We also recommend that in this mass program that is created where these this gain deltas are added that you maximize the noise reduction and directionality effects in order to minimize the effect of the loudness of environmental sounds around you and in order to maximize the speech to noise ratio because that is going to be the number one issue when using this program. Finally we also just you know suggest that you apply all best practices and audiological expertise to ensure that your clients will be you know their hearing will be safe and that they'll be satisfied with that fitting for these situations. So you know that that's what we ended up with those are the recommendations that we issued last year. They're on the Wet Nails website and with that I would like to thank you for your attention.