How would fitting software use Bone Conduction (BC) audiogram results?

I thought so, too, today, because in Bernafon Oasis you can enter that, too.

What it could do (I don´t know if it does): If there´s a big difference between air conduction and bone conduction, you have a disturbance of sound conduction. Then you can use more gain, because not all of the sound will get to the cochlea anyway.

If there is no difference between air and bone thresholds, all sound that comes out of the aid will get to the cochlea, so you must be careful not to hurt the residual hearing. You might even want to add more compression in this case, to compensate recruitment. But how bad recruitment really is depends on whether the outer hair cells are damaged or the inner, which cannot be guessed on the difference between air conduction and bone conduction.

So if you have the software, you might want to try (not while wearing the aids, it might hurt you): Enter the same audiogram for air conduction, once with bone-conduction equal, once with bone conduction 20 dB better. If I´m right, you will get more gain in the latter case, or at least less compression or a higher mpo.

@EnglishDispenser said:

BC losses are simply linear losses : x dBHL of loss requires XdB of gain

Sensorineural loss components (the AC value minus the BC value)are complex and nonlinear and require around 0.4db of gain per 1 dBHL.

Merci, monsieur, said the Yank to the Brit.

In my case, I have some BC thresholds that are 0-10dB above (visually on plot) / below (threshold value) the AC. Were the BC values omitted do you think there’d be any audibly noticeable difference in the fitting result?

Off-topic: if you click the link in EnglishDispenser’s signature, you’ll see a flame war thread from five years ago among English dispensers; being English, the flames are not as hot as they tend to be among American flame warriors. Also, the first post has a link to EnglishDispenser’s web site, which contains a wealth of information about hearing aids in general.

@EnglishDispenser said:

BC losses are simply linear losses : x dBHL of loss requires XdB of gain

Sensorineural loss components (the AC value minus the BC value)are complex and nonlinear and require around 0.4db of gain per 1 dBHL.

Could you elaborate further? BC loss: The loss that is shown by the bc-measurement is, as I understand, the sensorineural loss, because the eardrum and the following bones are “bypassed”.

AC value minus BC value (or better: BC value minus AC value, as BC value is usually higher) is the part of the loss that´s due to disturbance of sound conduction. Or did I confuse something?

@Musician_72 said:

… BC value is usually higher [than AC value] …

Do you mean higher – towards the top – on the audiogram plot, i.e., lower dB threshold values; or higher dB threshold values and lower on the plot?

In my case the BC values are at or above the AC values on the plot, i.e., having the same or lower dB threshold values.

Of course you are right - doh!

The BC value is the loss of the base nerve system.

The AC value is the base nerve loss PLUS the conductive loss.

So the conductive loss component is the AC value minus the BC value.

Sorry about the confusion!

@audiogal said:

I’d say they are if there’s a conductive loss, since you’re going to need more gain in that case. I don’t enter them for sensorineural losses, though.

The hearing aids that we fit, are they air conduction or bone conduction aids? They’re air conduction aids 99.9% of the time. Why would a 65dB conductive hearing loss need more gain then a 65dB SNHL if recruitment isn’t an issue? I have never been told by any hearing aid manufacturer that their fitting software calculates targets differently if there is an air bone gap vs. if there isn’t.

— Updated —

@SteveAUD said:

The hearing aids that we fit, are they air conduction or bone conduction aids? They’re air conduction aids 99.9% of the time. Why would a 65dB conductive hearing loss need more gain then a 65dB SNHL if recruitment isn’t an issue? I have never been told by any hearing aid manufacturer that their fitting software calculates targets differently if there is an air bone gap vs. if there isn’t .

Jeebus: you didn’t just write that, did you?

@Um_bongo said:

Jeebus: you didn’t just write that, did you?

Let me specify, it depends on the degree of the conductive component. 10dB air bone gaps vs. 50dB air bone gaps. I didn’t mean a 65dB conductive hearing loss, what I mean to say was a 65dB hearing loss with a conductive component (like one found often at 4k-falsely).

— Updated —

@brec said:

Can I infer that any audiogram variations, including my AC-BC gaps, not larger than 10dB may be merely random variation?

That’s correct. I’d only enter bone scores if there’s a significant difference from air (+10dB) since less than that won’t make much difference. Just went to a Siemens training seminar and they specifically mentioned that their software does take AB gaps into consideration in their calculations. I’d assume that most of the others do as well.

My audiometer spits the audio directly into Noah and therefore the fitting software for me, so I don’t have to worry about whether or not I want to enter information.

A colleague on qualifying had some spare time so he actually tested fitting software handling of B/C points entered on the audiogram.

He found that the fitting software handled the data VERY poorly. Many fitting programs wouldn’t even look at B/C data unless at least three points were entered.

Others required the B/C points to be entered on BOTH graphs (L+R) even if not masked even 'tho you only measure once.

Overall, he was VERY unimpressed with the way fitting software handles B/C data.

For conductive losses I suggest you check the calculated gains against [conductive component db loss + 40%*sensorineural component loss].
Change the gains by hand if there is a mismatch. Also increase the MPO to allow for extra power needed to punch through the conductive blockage.
You might also need to select a higher gain and higher power aid than that recommended by the fitting software.

Finally work with the client on a trial and error basis to make sure the settings are as good as you can get.

>> Can I infer that any audiogram variations, including my AC-BC gaps, not larger than 10dB may be merely random variation?

Yep. The audiometer may look like in came out a spaceship …but humans are simply blobs of analogue gloop, so all those fancy LED and screen displays may not actually be very precise or repeatable.

A calm and happy well-rested client may give good results whilst the same person in the evening after a noisy car journey in the rush hour and after a difficult day at work may have very different test results.