Re: Definitive tests of Bell resonance theory?

[Andrew Bell <bellring@xxxxxxxxxxxxxxxx>]

Dear Christopher and List:

Yes, these are important points that you raise, and deserve explanation.

>In their paper  "Is the pressure difference between the
>oval and round windows the effective acoustic stimulus for the cochlea?"
>(JASA 100:1602-1616), Voss et al. compare cochlear potential responses
>to common- and difference mode pressure inputs at the oval and round
windows.
>They find that the response to the pressure difference mode is
approximately 2
>orders of magnitude larger than the response to the common mode (with the
>limit set in large part by limits in their ability to measure extremely
small
>pressure differences).
>Naively, at least, this seems to contradict your model, which suggests that
the
>common pressure mode within the cochlea constitutes the effective stimulus
to
>the ear.

At first glance, these findings do appear counter to my theory, but you
should appreciate that my theory is most strongly focused on explaining
auditory behaviour below about 60 dB SPL, the region in which the cochlear
amplifier works. At these low levels, I propose that a surface acoustic wave
(SAW) resonator, comprising OHC1 and 3 acting in anti-phase with OHC2, acts
as a regenerative receiver of acoustic energy. A second aspect of my theory
addresses cochlear mechanisms above 60 dB; here sufficient differential
pressure develops across the tectorial membrane to generate ripples at the
vestibular lip, which again propagate along the underside of the TM to the
IHC, but this time in a different direction. (At 60 dB, cancellation of
ripples from both directions occurs, and this explains the notches and
peak-splitting in IHC responses that are commonly observed.)

The important thing to note is that the experiments of Voss et al. typically
involved high SPLs (90-120 dB in Fig. 5), and in this region I would agree
that differential pressure is the major stimulus for the partition (the SAW
resonator saturates at 60 dB when the intracellular potential of OHC2 falls
to the same as that of OHC1 and 3, and the cochlear amplifier runs out of
steam). I predict that if these authors had conducted their experiments
below 60 dB, they would have found a strong common-mode response. The only
data shown below 80 dB is in Fig. 4, and indeed the graphs show that very
strange things are starting to happen below this level.

Note that this paper did not rule out some common-mode response even at high
SPLs. The authors found that the difference signal, based on cochlear
microphonics, is about 35 dB greater than the common-mode signal. Although
this CMRR is large, the authors conclude that it is probably not infinite.

You also note my requirement for compressibility of OHCs (and hence the
cochlea) as an efficient means for funneling acoustic energy to the hair
cells, and relate that:

>In their ARO abstract (1996, #227), Ravicz et al. measure
>the compressibility of the cochlea. Their results are consistent with the
>known compressibility of water. Of course, these measurements were
performed
>in a temporal bone preparation, so their applicability to the living ear is
>uncertain. Whether or not they contradict your model depends on the
>mechanism by which the OHCs are supposed to compress and on what the
>value of their compressibility is predicted to be.

The actual abstract on the WWW reports a CMRR of 10-20 dB, which equates to
a compressibility higher than that of water, and is certainly not
inconsistent with a small compressibility figure. The compressibility I am
seeking does not need to be large: if we consider a piece of quartz, we may
regard it as incompressible by ordinary measures, but its small
compressibility is the basis of the important piezoelectric effects that
allow this material to be a useful pressure transducer.

I also take your own work in this area to be relevant. In your article 'An
empirical bound on the compressibility of the cochlea' (Shera and Zweig,
JASA 92, 1382-88), you derive a figure of 1% as giving the best fit to the
data. The paper points out that the figure could not be zero, otherwise
people lacking middle ears could not hear at all (yet they do), confirming
Bekesy's original investigations.

Additional evidence for compressibility comes from Kringlebotn (1995) [JASA
98, 192-196], who measured the volume flow between the round and oval
windows. Data for a human temporal bone give a volume displacement of the RW
some 3 dB less than that of the OW (much bigger than the equivalent Shera &
Zweig figure of 0.2 dB). Kringlebotn says that an earlier study gave a
comparable value (of 3 dB). Studies of 8 pig ears gave a figure close to
zero, but all the data points combined (Fig. 2) tend below zero; one data
point (1 kHz) shows a value of about 1 dB and the error bars do not include
zero. I therefore disagree with the author's statement that the volume flows
are equal when, on inspection of the data, they are not. Kringlebotn says
the discrepancy could have come from partial fixation of the stapes, but of
course if the cochlea contents were incompressible, this would have no
effect on the equality of volume flow between OW and RW at all.
Significantly, all the 'anomalies' tend to point to the volume flow through
the OW being higher than through the RW - just what one expects from
compressibility in the cochlea.

Direct evidence for compressibility of OHCs comes from microscopic
observations of these cells. Thus, when Zenner (1992) placed OHCs in a
tight-fitting pipette and applied a negative pressure of 98 Pa, the OHC
changed their length (and hence their volume) by about 2%. Iwasa and
Chadwick accepted this volume change [JASA v94 (1993), 1156-59], but
interpreted it as due to a (slow) flow of water and electrolytes. My
interpretation is that it is due to compressibility. It would have been a
valuable piece of data if Zenner had said how quickly the length changes
occurred. That he didn't state the time leads me to think it happened
quickly.

I suggest that the whole point of the layered construction of the OHC's
subsurface cisternas is to allow compression: presumably these layers are
filled with a non-aqueous compressible fluid. The other possible candidate
for compression is the globular Hensen body, which again has a lamellar
construction.

Other experiments have shown OHCs to be responsive to oscillating pressure.
The most pertinent have been those where mini-hydrophones (oscillating water
jets) have been directed at OHCs and their response observed. The classical
interpretation is that they are responding to deflection of their
stereocilia, but Brundin & Russell (1993) found that the response did not
depend on the presence of hair bundles and they suggested that "OHCs may
change length directly in response to sound". Canlon & Brundin (1991) found
that OHC length changes in response to their hydrophone "were not elicited
when the stereocilia were stimulated directly". The length changes observed
were tonic only because the applied stimuli were overloading the cells
(Canlon & Brundin were seeking a visible length change of 0.5 um; in vivo,
smaller oscillating length changes in phase with the stimuli would occur, a
similar mechanism to that which produces Brownell's electrically evoked
emissions). Note that the length changes observed by C&B had dual
polarities: lengthening (supposedly from OHC1&3) and shortening (from OHC2).

I therefore conclude that the question of compressibility, whether of the
cochlea in general or OHCs in particular, is still open, and that my
resonance theory of hearing remains a viable explanation of how the ear
works. Definitive tests (at levels below 60 dB SPL) have yet to be done.

A convincing demonstration could be done by performing hearing tests on
someone who lacked a middle-ear. I predict that this person's hearing
responses would saturate at levels corresponding to the limits of the
cochlear amplifier. A strong pointer to what is going on already exists in
Bekesy's 'Experiments in Hearing' (p.107); there he demonstrates that the
ear of a person without a middle ear showed a 180-degree phase difference to
the response in his remaining good ear. That is, I interpret the subject to
have a common-mode pressure response in one ear and a differential pressure
response in the other. Judging by path-lengths on a cross-sectional diagram
of the TM, the out-of-phase response is likely to derive from a 3-1/2 cycle
path difference at the IHC between the OHC (common-mode ripple source) and
the vestibular lip (differential-pressure ripple source). I would predict
that if the intensity of the sounds to both ears were lowered to where both
were responding to common-mode pressure alone, the phase difference would
disappear.

My personal position is that I would prefer confirmation to come from tests
on consenting subjects rather than from death or injury to animals.

Thankyou Christopher for pointing to these major considerations. I am
addressing them fully in the JASA submission, but I am always happy to
elaborate particular aspects.


Andrew.




-----Original Message-----
From: AUDITORY Research in Auditory Perception
[mailto:AUDITORY@LISTS.MCGILL.CA]On Behalf Of Christopher Shera
Sent: Tuesday, 11 July 2000 12:20
To: AUDITORY@LISTS.MCGILL.CA
Subject: Definitive tests of Bell resonance theory?


Andrew,

In describing your resonance theory you write...

> I have sought to ... develop a true
> resonance theory in which outer hair cells sense the common-mode pressure
in
> the cochlea. [The theory] calls on the body of the OHCs to respond to fast
> intracochlear pressure and in reacting (OHC2 in-phase to pressure; OHC1&3
in
> antiphase to pressure) to create a surface acoustic wave (SAW) resonator.
[...]
> [As a result]  it is the fast compressional wave that sets up a  pressure
regime
> in the cochlea which launches more slowly evolving wave activity.

To the extent that I understand all this, I believe that a definitive test
of
this
hypothesis has already been performed; the results appear to contradict your
model.

In their paper  "Is the pressure difference between the
oval and round windows the effective acoustic stimulus for the cochlea?"
(JASA 100:1602-1616), Voss et al. compare cochlear potential responses
to common- and difference mode pressure inputs at the oval and round
windows.
They find that the response to the pressure difference mode is approximately
2
orders
of magnitude larger than the response to the common mode (with the limit set
in large part by limits in their ability to measure extremely small pressure
differences).
Naively, at least, this seems to contradict your model, which suggests that
the
common pressure mode within the cochlea constitutes the effective stimulus
to
the ear.

You also write...

> If OHCs are pressure detectors, it follows that they must possess
> some compressibility.

In their ARO abstract (1996, #227), Ravicz et al. measure
the compressibility of the cochlea. Their results are consistent with the
known compressibility of water. Of course, these measurements were performed
in a temporal bone preparation, so their applicability to the living ear is
uncertain. Whether or not they contradict your model depends on the
mechanism by which the OHCs are supposed to compress and on what the
value of their compressibility is predicted to be.

Perhaps you can address these issues in your submission to JASA.

-Christopher Shera

--
Christopher Shera                               617-573-4235 voice
Eaton-Peabody Laboratory                        617-720-4408 fax
243 Charles Street, Boston, MA 02114-3096       http://epl.harvard.edu
"Sadism and farce are always inexplicably linked." -- Alexander Theroux