[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Cochlea Amplifier models : a new list

Hello Dick, Matt, and List,

Today, too, I shall reply to your points later, with the
exception of Matt's first question on "the point I am trying
to make". As probably many list members have guessed,
my main point is the presentation and discussion of a
specific cochlear model, which very probably is not new,
but which for shortness I shall now call "my model":

1) There are "BM resonators" (spring = BM fibres,
mass = organ of Corti).

2) There are "HC (Hensen-cell) resonators" (spring = OHC's
and maybe parts of Deiters cells, mass = Hensen cells and
other nearby structures; the RL pivots around the apex of
the pillar cells and thus varies the angle which it forms with
the BM).

3) At any given location on the BM, the resonance
frequency of the HC resonator is lower, by about one
octave, than that of the BM resonator.

4) x_b = distance, on BM, from base. During a soft or
medium-loud sine tone, the HC resonators basal of the
basal limit of the active-peak x_b-region are not excited
significantly, because their resonance-peak frequency
region is above the sine-tone frequency. The OHC's in
that basal x_b-region do not feed energy into the travelling
wave (TW) and thus are "passive". In a short x_b-region
above the just mentioned x_b limit, the sine-tone frequency
is within the resonance-peak frequency region of the HC
resonators; these resonators are excited, and the OHC's,
by their motor proteins, feed energy into the TW. At
locations with x_b above the active-peak centre, the
resonance frequency region of the HC resonators is
below the sine-tone frequency, and the OHC's feed no
energy into the TW.

Now I would like to encourage interested list members to
go to the library and look at some more figures in journals
which in my opinion are consistent with "my" model:

1) A. Fridberger and J. Boutet de Monvel (2003), "Sound-
induced differential motion within the hearing organ",
nature neuroscience 6, 446-448. Figs. 2b and 2d show
that the rotation centres of the guinea-pig cochlear-explant
RL and TM during 160-Hz sine-tones did not coincide in
space, but were separated from each other by about 0.05
mm. The RL rotation centre was close to the IHC's, and the
BM rotation centre was either at the feet of the inner pillar
cells or under the nerve fibres contacting the IHC's. The
observed place was near the 160-Hz characteristic place.
The sound was very loud  (126 dB SPL), but the effective
level was reduced by opening of the cochlea and immersion
of the preparation in tissue culture medium. Nevertheless,
the energy generated by the OHC's was, because of the
high level, probably not large in comparison with the energy
fed to the observed region on other paths.

2) P. J. Kolston (2000), "The importance of phase data and
model dimensionality to cochlear mechanics", Hearing
Research 145, 25-36. Fig. 9 shows results of Paul's finite-
element calculations giving a good fit to soft-sine-tone 9-kHz
chinchilla BM vibration level and phase data of Ruggero et al.
This Fig. 9 approximately confirms the earlier-mentioned
Fig. 3 of de Boer and Nuttall (1999). In this Fig. 9, the (only)
successful model is that represented by the dotted curves
("stimulus enhancement"). The real part is negative (OHC's
feed energy into the TW) from x_b = 1.2 mm to x_b =
3.6 mm; the active peak (Fig. 8A) is at x_b = 3.7 mm, and
the place where the imaginary part of the BM impedance
vanishes (i.e., the place of the BM-resonator which has a
resonance frequency of 9 kHz) is within the graph in this
case, namely at x_b = 5.0 mm. The reason why no passive
peak is visible in this Fig. 9 is given under point 3 below.

3) A. Recio et al. (1998), "Basilar-membrane reponses to
clicks at the base of the chinchilla cochlea", JASA 103,
1972-1989. Fig. 10 shows gain-versus-frequency spectra
for two animals. Here the frequency domain is used, i.e.,
a fixed place at about x_b = 3.7 mm is considered. Laser
velocimetry and Fourier transformation were used. The
sine-tone frequency in this Fig. 9 ranges from 3 to 11 kHz,
and the levels for the 9 living-animal curves were 26, 36, ... ,
106 dB for animal L13, and 24, 24, ... , 104 dB for animal
L113. In addition, there are two post-mortem curves at 96
and 106 dB (L13), and at 94 and 104 dB (L113). Animal
L13 has especially high gains. At 26, 36, 46, 56, and 66 dB,
the highest peak is at 9.0 kHz. This peak is absent in the
post-mortem curves and thus can be said to be the "active
peak". The post-mortem curves are about equal to the
106-dB and 104-dB curves and exhibit the broad passive
peak, at 6 kHz, i.e., about half an octave below the active
peak. The L13-curves from 76 dB to 96 dB have their
highest point at values between 6 and 9 kHz; that can
be understood by looking at Fig. 9, where BM velocity
(rather than gain) is shown: at 66 dB the OHC-Amplifiers
generate their maximal energy. At higher  levels, the height,
in dB, of the active peak is reduced, so that the active peak
degrades to a small spike on the high-frequency slope of
the passive peak. In Fig. 10, the gain is level-independent
below 6 kHz, because the resonance-peak frequency-region
of the local HC resonator is above 6 kHz. Thus for
frequencies below 6 kHz the TW is entirely passive from
the base to the considered place of x_b = 3.7 mm. In the
26-dB curve of Fig. 10 (L13), the passive peak is not clearly
visible because the low-frequency slope of the active peak
starts at 6 kHz. I hope to be able to discuss the additional
peak at 8.1 kHz in the L13-curves at 56 to 86 dB later.

4) A. R. Cody (1992), "Acoustic lesions in the mammalian
cochlea: Implications for the spatial distribution of the
'active process' ". The right half of Fig. 2 shows results
obtained with a living-guinea-pig cochlea with completely
destroyed OHC's from x_b = 2 mm to x_b = 4 mm. The
BM sensitivity for soft sine-tones of 23, 18 and 14 kHz
was strongly reduced, but that for 13 kHz was normal.
The characteristic place (i.e., the active peak) for soft
13-kHz sine-tones in guinea pigs is at x_b = 4.5 mm.
So, during a soft 13-kHz sine-tone, in a healthy guinea-pig
cochlea, only the OHC's located between x_b = 4.0 mm
and x_b = 4.5 mm are active.

Reinhart Frosch.

Reinhart Frosch,
Dr. phil. nat.,
r. PSI and ETH Zurich,
Sommerhaldenstr. 5B,
CH-5200 Brugg.
Phone: 0041 56 441 77 72.
Mobile: 0041 79 754 30 32.
E-mail: reinifrosch@xxxxxxxxxx .

----UrsprÃngliche Nachricht----
Von: DickLyon@xxxxxxx
Datum: 18.10.2007 23:48
An: <reinifrosch@xxxxxxxxxx>, <AUDITORY@xxxxxxxxxxxxxxx>
Betreff: Re: Cochlea Amplifier models : a new list

Matt says you are pointing out two distinct peaks.  So I
repeat my question from before:  does the data support
only that interpretation?  Or can the two (active and
passive) peaks be just as well interpreted as two ends of
a level-dependent continuum, reflecting a degree of active

Active traveling-wave models as I know them show a
continuous behavior between low-level active and high-
level passive, with degree of activity also influenced by
efferent control and local adaptation state.  Do you believe
the data support that notion?  Or contradict it?
Or neutral?