Will AI Hearing Aids be Boom or Bust?

Hearing Aids
#1

I think it’s still too early to make a call on AI hearing aids since we’re still in the early innings of development. With that said it appears Starkey was first out the blocks pumping AI HA technology. Maybe Oticon was second but I don’t think either aids have set the world of fire. Not sure when Phonak plans to jump on the band wagon but I’m sure in 2024 we see something from the Swiss company.

I guess the major question is will aid manufactures pump AI as a major jump in hearing aid technolgy/performance or just more bells and whistles that offer limited benefits? Hopefully in a year or two we start seeing new aids that reflect “major changes” in aid design and performance versus samo-samo.

4 Likes
#2

I’d guess AI will be minor refinement for what are devices with already really quite powerful signal processors. I note that Phonak’s Autosense is described as based on AI-trained machine learning and there’s improvement in every version but I think it will be hard to come up with a radical breakthrough.

1 Like
#3

No. Oticon was the first with the Oticon More in 2021. Then Phonak with using AI for the Autosense O.S.

Starkey have only recently have got into AI.

At the moment AI is here to stay. The senior director of Signia, Brian Taylor, talked up AI in a recent interview about the Signia IX, when he was asked about future developments - https://www.youtube.com/watch?v=DdjbV-bx8tM (about 20 minutes into the interview).

Potential applications, he says, are training the system to learn voices of interest, say, your mother, friends etc.

3 Likes
#4

I think DNN and AI will be “bells and whistles” because output of hearing aids is a sound - better or worse - so our broken cochleas must perceive it and our brain behind broken cochleas must process it. There will be no significant changes.

3 Likes
#5

Yeah, have to agree with this. At the moment AI seems to have had an incremental improvement, say, 5%, 10%, but nothing mind blowing. As you say, there is a loss in accuracy when sending a massively compressed signal to the inner ear, to cope with the reduced dynamic range in frequencies across the board, resulting in a loss of fidelity. Especially for severe/profound losses.

xMEMS (solid state receivers) might be a game changer, as this technology has demonstrated the potential to reach high frequencies with greater accuracy (whereas current hearing aids are unable to provide sufficient gain at all or cannot manage the feedback issues), as there are no mechanical parts… You might have seen this interview: https://www.youtube.com/watch?v=6C4Rft9G3TE

But the issue here could be feedback.

1 Like
#6

I have the Oticon More1 and they are light-years a head and other hearing aids i have ever worn.

4 Likes
#7

The claims for Oticon More on the Oticon website are as follows:

• ‌ Delivers 30% more sound to the brain**
• ‌ Increases speech understanding by 15%**
• ‌ Reduces listening effort by an astounding 30% over time***

So, it’s a little more than 5% or 10%. In terms of speech understanding.

Of course, the effect of more sound to the brain cannot be understated - hearing more soft sounds perhaps.

Whether these improvements are all down to AI or not, I don’t know, as there are other factors, such as faster chips/hardware platform.

#8

I don’t care what the specs say ,I know how much better i am hearing.

3 Likes
#9

I have profound hearing loss. My hearing aids in last 10-15 years was Widex Senso P38, Phonak Naida V UP (Core), Oticon Chili SP9, Oticon Dynamo SP8, Oticon Xceed SP 2, Oticon Xceed SP 1. All of these hearing aids gave me about the same percentage of speech recognition. So cochlea and brain is a main limiting factor - no hearing aid. Hearing aids can process speech in noise - they simply enhance and improve the sound. Yes, xMEMS will improve high frequencies for profound hearing loss (and will reduce the size of hearing aids) - but this will not improve brain. 15-30% from Oticon is a marketing.

4 Likes
#10

Yes. I think the claims made by whatever manufacturer, whether it be Resound, Phonak, whoever, do not say which segment the claim applies to. A 15% improvement for whom? I remember many years ago that in the small print, Siemens pointed out that the improvements in SNR for a particular aid (I think Binax) were only applicable to those with moderate loss. But when I see similar claims now, they do not categorise who these claims belong to. My understanding from Oticon “brainhearing” is they make use of speech cues. But if these speech cues fall in the high frequencies and only end up being perceived to those with say, moderate to severe loss, and not severe to profound, then these improvements are not going to apply for the latter group.

There is insufficient transparency in this area, unfortunately. The manufacturers want to sell hearing aids, and to a large extent, the claims may be valid for people with mild/moderate - even slight severe loss, but for more severe losses, they may not be.

4 Likes
#11

We all have different hearing loss and different hearing needs. The whole hearing aid process is a crspshoot. I have l been lucky enough to find aids that work the best for me and i have stayed with Oticon for 13 years now. I will be getting the Oticon Real1 aids this next Thursday, which will make my 6th set of Oticon aids. I get my aids through the Veterans Administration due to my hearing being classified as military related loss.

1 Like
#12

@cvkemp , I will be very interested in your impressions of the Real vs. the More. Hopefully we can look forward to a detailed comparison report. I would very much appreciate it.

1 Like
#13

I can only speak for my experience with Starkey. Something like full 2-weeks trial with a Ai integrated RIC model. To me, a couple of things stood out.

  1. The ability to get an adjusted setting via apps is there. At least for the Thrive app, I think it can impact up to 10-15%, playing with “some” background noise, microphone directionality and similar regulation. Sometimes HAs disconnect from the app, it happens quite frequently. The new genesis AI have a new brand exclusive app, so I can’t say if there will be substantial changes. Keeping in mind that I’m not sure what role the AI play – on the Thrive app specifically there was an “AI optimal setting” option and irl for me it never aced the setting based on the environment I was in. Honestly I think that was because the setting I had was pushing to the max high frequencies, therefore there was a lot of whistling – in other words, I’m not an ideal trial

  2. I add a specific “complain” regarding Starkey and the innovative AI direction that has been taken. Premise. I do understand how even minimal design choices (like very good improvements on dropping weight) and AI implementation require a lot of research, money and time to get developed and marketed. Ok, fine.
    From a user point of view, there are some features I can’t rely on ( (transcription is very weak) or that I don’t use that much (health measures are not related to better hearing devices in my opinion) that were prompted at the expenses of Bluetooth. More in detail I don’t know if I’m the only one to consider Bluetooth the biggest improvement made for HAs in the decade 2010-2020. I use it everyday. So, I make this simple consideration: the Bluetooth technology in Starkey is a 2.4Hz system and it never got implemented in over 7 years. Why I say that? All series of Bluetooth accessories to connect to a PC (for now I can’t connect directly via HAs), table microphone etc are the exact same there were 7 years ago, maybe even prior. If I’m not proven wrong, for the newest Genesis AI line there are not updates Bluetooth accessories. Is this innovation? Even the most powerful HAs or CI require external aid for a complex hearing environment and I don’t know if Ai or Bluetooth are intended to work in sync or else.

1 Like
#14

HALLELUJIA AND HOT DANG! That is what’s been on my “wish” list for decades. Just give me the SPEECH. Based on learning or some kind of smart frequency processing. Human speech is on a spectrum, but I still want it somehow separated out and boosted from say, crashing plates at a restaurant, LOUD indistinct blather at parties, etc.,

HA makers promote AI used to detect if one falls down, or what their heart rate is, but for me? I want human speech comprehension above and beyond what I’m even getting in the Phonak Lumity Life aids. And that said, I continue to be amazed at how much better I comprehend speech even in challenging places (restaurants, high-ceiling’d places, etc.,) than with my previous generation Marvels.

Speech comprehension gives confidence. And that begets INDEPENDENCE! I sometimes even tell my husband what’s been said these days. My comprehension has improved that much in the past year. :smiling_face:

4 Likes
#15

I also feel that hearing 5%, 10%, 15% or even more is all a positive move in the right direction! The brain-ear connection is like a muscle that can be strengthened and improved similar to how our biceps can be trained to lift 5-lb, 10-lb and MORE weights.

I was “sold” on the so-called 10% speech comprehension improvement touted by Phonak for their Lumity Life aids. But somehow, that 10% improvement seems to have grown to about 18-20% in a year. I’m guessing that as I comprehend speech more and more, that brain-ear connection sharpens up so I don’t have to hunker down and CONCENTRATE to make out the words. It’s so liberating to be able to participate in even fast interchanges, making out what’s been said.

Now, if only AI would zero in on human speech frequency even further - or perhaps identify speech (based on words, pauses between them, tone, inflection) versus ambient sounds in order to BOOST that. I think it can be done and WILL be done for aids in the coming 5 yrs or so?

3 Likes
#16

Just want them to work when I put them in and turn them on. 20 years of hearing aid use…

FWIW my Phonak Audeo Paradise P90R’s didn’t work for over two years.

Let’s hope that future AI can keep an ear on what our providers setup. It should be simple to set them up.

It should be specially simple if a Phonak hearing aid is replacing previous Phonak hearing aids.

It should be easy to get hearing aids that work right from the beginning. My experience…oh my.

DaveL

#17

I have been wearing Oticon aids for 13+ years. My audiologists have been useing the previous aids settings to program my next generation of aids all along. But to find that point took monthly appointments over about an 18 month period of time. I was told by my early audiologist that my hearing loss was impossible to really adjust to and to just accept what i could get. The audiologist i have now, for the last 5+ years has been the one to not accept good enough.

1 Like
#18

@cvkemp cvkemp

My old auddi always had a reason why I couldn’t hear. 2 years.

DaveL

1 Like
#19

The audiologist i have now gave me his email address and private phone number, and said to contact him directly if i was having issues

3 Likes
#20

Please beware that AI is often used as a marketing term and that there is no real definition of AI afaik.

3 Likes