New? "Neuro- Compensator" Hearing Aids

On this forum some new manufacturer or sales outfit is touting a new aid that features a “Neuro-Compensator”.

I read the web site description and it is replete with a gobbley-gook of technical terms that seem to me to be a restatement of what all modern aids do.

I wonder if any of the pro’s on this forum have any knowledge of the validity of the product. Frankly, it sounds to this engineer like baloney. If I am mistaken please educate me. Ed :rolleyes:

Hello Ed,

The Neuro-Compensator hearing aid is commercialized by Vitasound Audio through several clinics in Canada and the US. It’s a truly unique approach to hearing aid amplification, relying on a computer model of the patient’s auditory system to optimize the amplification parameters: NO compression at all, but a rather complex gain formula that has been developed at McMaster University (Hamilton, ON, Canada).

Each time we introduce the NC to a new partner/clinic, we typically offer a 30-minutes online session which I will be glad to introduce to you at your earliest convenience, if you wish. In the meantime, please, do not hesitate to contact me if you have questions.

Philippe A. Pango, Ph.D
Chief Technology Officer
Vitasound Audio Inc.

Yes, how does a retired working stiff decipher tech jargon and sales hype from true value for improving hearing loss. I have been struggling with this for years.
It seems the only way is to rely on experience (costly) and sound advice which is hard to come by.
I going this morning to have my hearing tested and discuss replacing my 8 year old Widex Diva’s.
And the battle continues.

Sounds like a different way to do the same old thing as every other hearing aid to me. Could be wrong…but I don’t think so. Still a Hearing aid. Still amplifies sound.

Could you post more complete technical detail. Frankly it sounds no different than any brand of NTSC except the lack of compressive control.

I am not aware of any revolutionary hearing aid development that fits the description in the technical journals (Proceedings of the IEEE or the Journal of the Acoustical Society of America). Could you sight a reference? Ed

This is an interesting subject and I’ll try to give a simplified explanation that should be accessible to someone with an engineering or other technical background of how the “ultimate” hearing aid might work if such a thing existed!

It is believed that sensorineural hearing loss is due to hair cells in the cochlea no longer functioning properly. This leads to a soft sound at a frequency making it to your cochlea, but the signal not being passed any further (whereas in someone with normal hearing that sound would be passed on to the brain).

For someone with normal hearing you can imagine a certain level of input being perceived as an equivalent level of loudness. For example, an input of “30” would be interpreted as a loudness of “30” in their brain, an input of “50” would be intrepeted as a loudness of “50” in their brain, etc. In this example input would be some measure of the sound as it’s about to enter your ear (say dB HL) and loudness would be your psychological perception of it.

For some particular particular with a sensorineural hearing loss inputs and perceived loudness might map something like this at a given frequency:
20 -> 0
30 -> 0
50 -> 5
55 -> 30
60 -> 50
70 -> 65
80 -> 80
90 -> 90

As you can see, soft sounds aren’t heard at all below a certain level (a “threshold”), but as soon as that level is reached loudness scales up very quickly and at some point above that level you experience sounds just as loud as someone with normal hearing would.

Now let’s suppose you have a device that allows you to accurately measure perceived loudness in someone else’s brain. In theory you could create a hearing aid that would adjust sounds entering the ear to the exact level needed to achieve the desired loudness. For example (using the mapping above) the hearing aid would need to convert an input of 5 to a level of 50 before sending it into your ear. An input of 10 might be converted to 51. An input of 20 might be converted to 53. An input of 50 might be converted to 60. You get the idea…

So what stands in the way of us creating this “ultimate” hearing aid? Well, for starters the required mapping would vary frequency to frequency so we’d need an incredibly large number of bands (and processing power well beyond what’s available in hearing aids today) to accomplish this effectively. The “testing” procedure each patient would go through would be very extensive and would require equipment capable of scanning a patient’s brain and accurately detecting perceived loudness as tones of varying intensity are presented to the patient (I’m not aware whether or not this technology exists). A more sophisticated form of Real Ear Measurement would be required since very slight variations in input drastically affect perceived loudness.

I should point out that I’m leaving out a lot of other considerations that would need to come into play to avoid making this description too much more complicated than necessary.

This “ultimate” hearing aid would live up to their claim of “The Neuro-Compensator™ amplifies the audio bands so as to target a near-normal neuronal activity in the auditory system” Without utilizing the system I’ve described (which won’t be available in the near-future) you can’t really “target a near-normal neuronal activity in the auditory system” significantly more effectively than hearing aids currently available do.

And my next patient is here. Hope you enjoyed a glimpse into what the ultimate hearing aid would be like!

For someone with normal hearing you can imagine a certain level of input being perceived as an equivalent level of loudness. For example, an input of “30” would be interpreted as a loudness of “30” in their brain, an input of “50” would be intrepeted as a loudness of “50” in their brain, etc. In this example input would be some measure of the sound as it’s about to enter your ear (say dB HL) and loudness would be your psychological perception of it.

I’m sorry…and correct me if I’m wrong/misinterpreting your post…but what you described sounds exactly like an audiogram. Ultimately it doesn’t matter what level is required to stimulate part of the brain…if your brain doesn’t interpret it as a sound it doesn’t matter.

We are limited by what can be transmitted and received and interpreted by the brain. The only way to measure that full cycle is via behavioral testing which is typically…the audiogram. This nonsense about this neuro-whatever is just another way to package standard amplification principals but utilizing a different formula to calculate them. There was no mention of any new-fangled awesome way to amplify sound at AAA this year…not a peep. If this was some amazing new way to process/amplify sound so that it would do the amazing things they say it will, it would have been all over the place.

An audiogram plots the thresholds at which a person just begins to be able to hear sound. What I’m referring is like a loudness scaling contour. Imagine if you started with an audiogram, but instead of only plotting the thresholds you then plotted lines indicated how loud of a stimulus was required (at each frequency) for a patient to psychologically hear a loudness of 10 dB, 15 dB, 20 dB, etc. Again, for a patient with normal hearing a 50 dB sound would have a loudness of 50 dB, whereas for a patient with a hearing loss his perceived loudness could be less than 50 dB at a given frequency.

I agree. I think the “neuro-compensator” is marketing fluff. I was just pointing out that if they were truly going to achieve their claim they’d have to come up with a system similar to the one I described, which although theoretically possible does not currently exist.

Let me try describing this theoretical “ultimate” hearing aid from another angle:

Think of the pinna, ear canal, eardrum, ossicles, and the hair cells within the cochlea as a “thing” with the ultimately function of delivering an electrical signal to your eighth cranial nerve (and ultimately your brain).

Now imagine if you could somehow monitor the auditory signal as it’s being passed from the cochlea to the eighth cranial nerve. That signal would be very different for someone with normal hearing VS someone with a hearing loss. If you tried to create a hearing aid with the goal of eliminating that difference and making the two signals match that would be a huge step beyond what hearing aids do today. Easy to say, but no so easy to do.

That seems to be what VitaSound is implying that their hearing aids attempt to do. The only problem is that absent all of the additional steps and technological advances described above I don’t think it’s possible to pull something like this off!

Or simply, REM is often done to better match a hearing aid’s output to the sound level that actually exists right next to a patient’s eardrum. The “ultimate hearing aid” would start by matching a hearing aid’s output to the “sound level” that is actually being passed to the brain.

OOoohhhhh…okay. That makes more sense to me…I see where you are going with that.

Folks,

I will do my best to explain what Neuro-Compensation is, in extremely simple terms, hoping this would allow you to realize how different this approach is to ANY existing hearing aid out there.

Let’s start with a simple analogy. Eyeglasses distort the incoming light before it reaches the eye and thus, it is the combination of an altered light mixed with an eye that doesn’t function properly which restores sight, correct ? Similarly, the purpose of a hearing aid is to distort all sounds (particularly speech) such that, the combination of this distorted sound with a bunch of hair cells that don’t operate properly, “fools” the brain by (hopefully) sending the appropriate signal through the auditory nerve.

That being said, everything we hear, everything we see, everything we smell, touch or taste ends up being converted into a series of tiny little electrical pulses that travel really fast through our nerves and reach the brain, our ultimate processor. Are you still following me ? ok.

Now, imagine for a minute that your dispenser is equipped with a piece of software that allows him/her to visualize how YOUR damaged hearing system CONVERTS SOUNDS INTO ELECTRICITY (the neural response), the series of pulses that your auditory nerve sends to your brain ? In addition, imagine for a minute that your dispenser is capable of seeing how each of his fitting decisions affects the actual nerve output, and how that compares to the neural response that would be generated by an average healthy auditory nerve under the same acoustic stimulation ?

THAT’s what the Neuro-compensator is about:

  1. Modeling the patient’s actual response to sounds
  2. Figure out how the aid should alter the incoming audio, such that by cascading it with the patient’s ear, the patient’s auditory system would still generate an electrical response as close as possible to the one an average healthy ear would send to the brain.

Finally, 99.9% of all hearing aids out there use compression because the ENTIRE industry is stuck in that amplificaiton scheme which dates back from the analog years. Despite the introduction of digital technology in 1996, noone has been able to offer a truly NEW amplification scheme. Yes, compression provide some benefit, but only up to a certain point; that’s why 17% of hearing aids owners don’t use them.

The Neuro-Compensator does not use compression; it process the audio signal as a whole, even allowing the energy at one frequency/band to affect the gain dynamically applied at other frequencies/bands.

(Noe the geeky part - Sorry, I am an engineer)
The auditory model we use has been developed over decades by eminents Professors such Prof. Sue Becker and Ian Bruce at McMaster University (ON, Canada). The final Neuro-Compensator hearing aids took years to develop. Some of the finest minds in the hearing aid industry have been involved in its development (Algorithm, microchip design, etc). We have absolutely no doubt we are introducing a unique technology that truly helps people and enhance their quality of life.

Philippe
PS: We were not at AAA because our growth strategy did not require us to attend AAA. We are busy expanding our distribution network in Canada and the US.

Setting any audiological issues aside:

  1. The product name is very very silly - your marketing people should be executed at dawn

  2. The website is poor - very poor.

  3. You lack marketing focus : why do you link to an earphone supplier and why have you made a weird Personal Amplifier? And hearing protection too? And that Instant Custom Fit page? You might as well start a car rental branch or a burger bar.

  4. Your only key asset is IPR (the patented idea) so why bother with making STUFF? Why not licensed your super-duper new scheme to a real manufacturer? Unitron is in your state - why not give them a call?

  5. You should attend AAA. No brainer.

  6. NOAH compatibility might be a wise move seeing as maybe 90%+ of real dispensers use it.

  7. Your software is available to download - but then a registration code is needed. You could mention this requirement BEFORE I fill up my PC with your fitting stuff … and 800+ MB of MS Net Framework 4!

  8. You need registration for your newletter. Just make it available openly guys.

Your investors need to carry out an urgent project, product & marketing review.

I’d love to know what your corporate exit strategy is! Or if you have any corporate strategies at all …

And your ‘elevator pitch’ must be a sight to behold …

(BTW I have a corporate high-tech product background so I think I know what I’m talking about)

There are still more many variants

Erm.

Bernafon - Already sample and adjust at 9,600 points.

Channel-Free has been out for around 10 years.

1 Yes, but that’s like arguing you can wade across a river without getting wet and dirty, just by getting wet. You still have to correct the impedance mismatch generated by the dysfunction of the auditory system, the fact that you are visualising the output generated in terms of the nerve responses is a bit of a mechanistic answer - especially as even Brain surgeons use observable responses in ‘awake’ operations to guide their performance. If the machine doesn’t sound right, it’s a bit irrelevant as to how or why you amplify the noise.

Phonak have also claimed their aids are mapped to reflect the actual performance of the cochlear.

  1. Can you differentiate this from the Bernafon Channel Free system or any other energy averaging method that the other manufacturers use?

  2. Compression deals specifically with the recruitment of the damaged cochlear - how specifically can you introduce a linear model into a reduced dynamic range without compromising the edges? - both Adro and Oticon’s floating point linearty are methods of achieving this.

The complexity of how you’re adjusting each point comes into play here. If you’re simply adding a set gain value across different ranges that’s easier for the processor to handle at that many points. If you’re using a more complex “gain” function at each point significantly more processing power is required.

With that said, a modern computer (like the one I’m typing on now) could easily keep up, but I’m not sure whether or not that applies to processors currently being used in hearing aids. Maybe they can and I’m underestimating their current capabilities.

How exactly is the dispenser or your software monitoring the actual nerve output in the patients they’re working with? Short of sticking electrodes in the patient’s head or an (if one even exists for this) elaborate imaging system I don’t see how this could be achieved.

I still question the science behind it all. Let’s just say that the hearing aid is doing exactly what the designers are purporting it does, it doesn’t mean that the brain is making sense of the signal the way the hearing aid is intending for it to be interpreted. Plus, does the ear really NEED to have that level of specificity in amplification? Research has already shown that much above 10 or 12 channels (some would even say less) doesn’t provide any additional benefit to the patient. Also, unless you are stimulating the auditory nerve directly (like a cochlear implant which sounds nothing like normal hearing), you still rely on a damaged system to send the signal to the brain. That damage and the necessary amplification to bypass the damaged hair cells distorts the signal in addition to any (albeit small) distortion the hearing aid transmits.

EnglishDispenser, I really (and I mean really) appreciate ALL your comments. It is always a good thing to have such comments from someone out of the organization. Regarding the software, we typically do not expect people to download it just like that. Typically, emails are first sent to the potential distributor (with the registration key) before the download even takes place. Immediately after that, we provide an online training, because you can imagine our software operates differently. Indeed, because of the complexity of our fitting formula, an average computer takes more up to 4 minutes of processing to calculate the amplification parameters. For that reason, the Neuro-Compensator fitting software UPLOADS the patient’s audiogram to our server in Canada, and 4 minutes later, DOWNLOADS all calculated data back into the office, and into the aids. This is a unique process that requires an Internet connection at the point of fitting. That’s also the main reason a registration key is required, to prevent unauthorized access to our servers. Once the dispenser is familiar with this rather unusual process, fitting a Neuro-Compensator becomes a piece of cake.

The NC is a truly disruptive technology and its commercialization (we believe) has to go through a disruptive path too. We have been able to grow significantly while operating beside other manufacturers, and you have no idea how much interest this new technology is generating.

Again, thank you a million times for your remarks.

Of course we don’t stick electrodes in people’s head. We build a computer model of the patient’s hearing system, and then, the computer simulates this model while calculating the amplification parameters. It’s a rather complicated process using “machine learning algorithms” to optimize the fitting.

I know the paper below is VERY technical, but reading only the introduction and Section 2 (on the Auditory Model) should give you a very good idea of how science has evolved over the last decade.

Philippe