non-linearity IN BASILAR MEMBRANE AND OHC

Submitted to Sanjana mam

Assistant Professor

Department of Audiology , JSSISH

Submitted by,

G Jayanth

1st year PG

[Link]. Audiology
 1|Page

COCHLEAR NONLINEARITY

What is a linear and nonlinear system?

The auditory system can be thought of as a series of stages, the output, of a

given stage forming the input to the next. Each stage can be considered as a
device or a system, with an input and an output. For a system to be linear,
certain relationships between the input and output must hold true must fulfil
several criteria in order to be regarded to function as a linear
system. The output must increase in the same proportion as
the input is increased, and if two different input signals (such as
2 tones with different frequencies) are applied to the input of a
system, the output must be the sum of the output of the 2
signals when applied independently. This is known as the
superposition criteria of a linear system.
The following conditions must be satisfied:

1. Homogeneity:

If the input to the system is changed in magnitude by a factor K, then the output also
change in magnitude by a factor K, but be otherwise unaltered. This condition is called
Homogeneity. For eg. If the input is doubled the output is doubled.

Thus a plot of the output as a function of input would be a straight line passing through
the origin (zero i/p gives zero o/p). hence the term linear system. Such a plot is called i/p-
o/p function

When describing a system as linear, it is usually assumed that the system is time-
invariant. This means that the i/p-o/p function does not change over time. For eg: if the
input is I and output is O, relationship between them would be

O = CI , where C is a constant that does not vary with time.

2. Superposition:

The output of the system in response to a number of independent inputs presented

simultaneously should be equal to the sum of the outputs that would have been obtained
if each input were presented alone. This condition is known as Superposition.
 2|Page

Eg: if each input were presented alone; if the response to i/p A is X and response to i/p B
is Y, then the response to A and B is called as superposition

It is usually assumed that the linear system is time invariant. This means that the input
output function does not change over time.

The amplitude & phase of the output may however be different from those of the input ie
it Changes (or does not changes) only the amplitude and phases of input sinusoids.

The output of a linear system never contains frequency components that were not present
in the input signal. The response of a linear system may, however, vary with the
frequency of the input sinusoid.

It is relatively easy to predict its output in response to any arbitrary complex input.
Systems are linear as long as their input is not large. Excessive inputs may cause the
system to become nonlinear.

Nonlinear System

When sinusoids are found in the o/p of the

auditory system that are not present in i/p i.e. what comes out
is more than what went into the system. A simple waveform
enters the nonlinear system and a complex waveform comes
out the other end. The o/p may differ from the i/p in time
domain as well as in the frequency domain.

 The input-output function is linear over a large part of its range, but it flattens out
for large negative or large positive values of input. This is sometimes called saturation.
 The output contains components that were not present in the input. The output of a
linear system in response to a single sinusoid at the input contains one or more
components (sinusoids) with frequencies that are integer multiples of the frequency of the
input. These components are referred to as harmonics, and the nonlinear system is said to
introduce Harmonic distortion.
 If the input to the nonlinear system consists of two sinusoids, then the output may
contain components with frequencies corresponding to the sum & difference of the two
input frequencies, and the harmonics. These extra components are said to result from
intermodulation distortion. In case of an auditory system, they are also called
combination tones.
 3|Page

 The response to complex inputs cannot generally be predicted from the responses to
complex inputs. Waveforms are also changed in time domain by nonlinear devices. So a
system that adds sinusoids to input waveform is called a nonlinear system. When the
response to a sine wave is not a sine wave, but is still periodic.
 In case, the slope of input-output function is two, the nonlinearity is referred to as
Expansive. If the output were proportional to the square root of the input, the slope of the
function would be 0.5. When the slope of the input output function is less than one, the
nonlinearity is referred to as Compressive.
 Non-linearity in a sound system produces unwanted harmonic and intermodulation
distortion, and perfect linearity is the ideal. But for the ear, non-linear behavior is far
from being a flaw; in fact it is a critical feature that enables the large dynamic range of
human hearing.

The inner hair cells (IHC) of the cochlea, which convert sound to nerve impulses,
have a dynamic range of less than 50 dB. But we can hear over a 120 dB dynamic
range! How is this possible?

 It turns out that the ears have a built-in sound level compression system, created by
the outer hair cells (OHC) of the cochlea. In the most active region of the cochlea basilar
membrane, a 4 dB increase in sound pressure at the eardrum increases the membrane
motion as little as 1 dB, due to mechanical action of the OHC.
 The non-linearity of the ear has been known over a century, but it was relatively
recent that the OHC of the cochlea were identified as the primary cause. The middle ear
is quite linear over sound pressures of 40 to 110 dB SPL, and does not result in
noticeable distortion at normal listening levels. The inner ear non-linearity does produce
distortion, which can be heard, and measured in the ear canal.

Historical aspects

Helmholtz (1863) was the first to formulate and prove that ear performs spectral analysis
of sounds. Before that Ohms (1843) proposed that ear could separate a sound into its
frequency components. The BM functioned as series of resonators that were tunes to
different frequencies covering the audible range, a function similar to that of strings of
piano.
 4|Page

Von Bekesy (1942) showed that at certain point alone the BM, the vibration amplitude
has its maximal value at one specific stimulus frequency, & declines gradually when the
stimulus is increased or decreased. Frequency scale is laid along the cochlea with high
frequencies located at the base and low frequencies towards the apex of the cochlea.

By beginning of 1970s, it became evident that motion of BM is nonlinear. Rhode

(1971) showed that for single tone stimulation, the BM behaves nonlinearly as the
function of i/p level; the response is compressed for stronger stimulus. He also
proved that after death, the response becomes linear, indicating that the cochlear
functioning depends on the physiological state of the cochlea.

Dallos & Harris(1987) suggested that there may be frequency dependent boost in
the IHC sensitivity contributed by the OHC; they considered that OHC effect on IHC
may arise from electrical interaction via field currents or from an electromechanical
interaction.

COCHLEAR NON-LINEARITY

Non-linearity is an essential component of cochlear function and of OAEs. It has

already been mentioned in relation to sensory cell transduction and to the
compressive growth of TEOAEs with stimulus intensity. Non-linearity affects all
aspects of hearing and all cochlear responses

Cochlear nonlinearity is revealed in five main ways:

1) The frequency –dependent basilar membrane nonlinearities - The

nonlinear growth of BM responses with stimulus intensity
2) The frequency –dependent receptor potential nonlinearities – the
frequency dependent response level compression in the IHCs receptor
potentials.
3) The frequency-dependent two-tone suppression – the reduction in the
response to one stimulus by a second stimulus.
4) Distortion components or combination tone generated by the cochlea -
When two or more tones are presented simultaneously, humans can hear
additional tones that are not present in the acoustic stimulus.
5) The upward spread of masking - the suppression of a basal activator
stimulus (high frequency) due to the simultaneous presentation of an apical (low
frequency) suppressor.
5|Page

1) Basilar membrane nonlinearities:

 One significant aspect of the auditory system is its ability to process
sound over a wide range of levels, about 100-120 dB.
 The auditory systems large dynamic range and exquisite frequency
selectivity, both of which form vital components of our ability to
communicate in the acoustic environment, are largely established by the
mechano-electrical transduction that takes place in the cochlea. In
particular, the healthy cochlea responds to sound in a way that is highly
nonlinear and compressive.
 The mechanics of the cochlea begin with the vibratory input of energy
from the stapes, which in turn causes the oval window membrane to
vibrate in response to sound energy that is being transferred from the
middle ear to the inner ear. There are a variety of complex interactions in
the cochlea that involve the cochlear fluids, various membranes within
the cochlea and the hair cells that combine to define cochlear mechanics.
 The entire cochlear operation is highly dependent on the mechanical
properties of the basilar membrane (BM) and these mechanical properties
vary continuously along the entire length of this membrane. And the key
process in the mechanics of the cochlea is the generation of the travelling
wave.
 The tectorial membrane, which comes in contact with the OHCs, bends
the cilia as the hair cells "ride the movement of the BM. As the intensity
of the acoustic stimulus increases, the peak displacement of BM becomes
broader and grows in amplitude, but with considerable compression
across the Intensity range. This is especially the case at high frequencies.
 Ruggero (1992) did a study and concluded that these compressions that
resulted in non linearities in the displacement of the basilar membrane as
a function of intensity were measured at high frequencies.
 Because of the anatomy of the mammalian cochlea, it is difficult to
measure BM vibrations at the apical end; hence, the data on these effects
for low frequency stimuli are sparse. However, it seems that there is less
compression for low frequency than there is for high frequencies (Rhode
and Cooper, 1996). At very low intensities, the BM gets a boost from the
cochlear amplifier, resulting in increased BM excursions, although these
effects for low-frequency stimuli (ie., the apical end of the BM) are not
nearly as great as those noted for the higher frequencies (represented at
the basal end of the BM). At the higher intensities, some compression
6|Page

does not take place at the apical end of the BM, but again the extent of
these compression effects is not equivalent to those manifested at the high
frequencies. Finally, it should be noted that this phenomenon is not
limited to the CF (such as with high frequencies), but rather it is present
across most of the frequencies surrounding the CF.
 Therefore it can be concluded that BM excursions increase with
increases in intensity of the stimulus. However, the particular manner in
which the BM responds to increases in intensity is complex (Sellick,
Patuzzi, and Johnstone 1982; Ruggero, 1992; Ruggero, Recio, Narayan,
and Robles, 1997). Nonlinear compression takes place at high intensities,
and the area of maximum deflection of the BM becomes larger at higher
intensities, reducing frequency selectivity. Also, at high intensities the
cochlear amplifier is not active.

Two types of plots are used to study BM responses

1. The frequency response may be described an amplitude response, in
which the amplitude of vibration is shown as function of stimulus
frequency, the stimulus intensity being kept constant.

2. A tuning curve or frequency –threshold curve maybe measured, in which

the intensity of stimulus is adjusted so as to keep the amplitude of BM
vibration at a constant value, while the stimulus frequency is varied.

Fig: Amplitude Response curve:

7|Page

Fig: Tuning curve/ frequency –threhold curve/ isoresponse curve

To understand the compressive capability of the basilar membrane, it is

necessary to consider the increase in the magnitude of the basilar membrane
vibrations at a given point along the membrane as a function of stimulus level.

1) Rhode (1971), observed the basilar membrane vibration in squirrel

monkeys using Mossbauer technique. He revealed that a significant
degree of compression is already achieved in the cochlea at the
mechanical vibration stage. Measurements were made in the squirrel
monkey cochlea at the place corresponding to a characteristic frequency
of 7 kHz. Stimuli were presented at 70, 80, and 90 dBSPL.
The results of Rhode’s study differed remarkably from those that Von Bekesy
reported cadaver ears.
 The tuning of membrane was much sharper in the living ear. The
slopes of the tuning curves reached about 24 dB per octave just below CF
and nearly 100 dB per octave on high frequency side of the peak.
 The sharpness of the tuning curves increased as the stimulus intensity is
decreased. That is, as the stimulus level was reduced from 90 to 70 dB
SPL, the selectivity of the BM responses increased.
 Rhode called this non- linear phenomenon as “Negative Damping”. In
case of BM Negative damping really means “Amplification”. Outer hair
cells amplify the BM response to weak signals.
 8|Page

Cochlear amplifier and nonlinearity:

Although the term cochlear amplifier was apparently coined by Davis in 1983,
Cochlear amplification was first proposed by Gold in 1948

Much of the processing within the cochlea is dominated by so called cochlear

amplifier which is a hypothesized, nonlinear mechanism that acts simultaneously to
assist the filtering process & to compress the vibration of the BM over most i/p
amplitude range.

It is a positive feedback loop within the cochlea that amplifies the travelling wave.
If the loop is broken at any point, the BM is driven only by input stimulus.

 OHC motility can influence the tuning properties of the BM.

 The cochlear amplifier, which is mediated by the OHC function,
sharpens the frequency tuning, especially at low intensities. The phrase
cochlear amplifier refers to the properties of cochlea that are
physiologically vulnerable and leads to high sensitivity and sharp tuning..
 This loop, A-B-C-D-A  cochlear amplifier. Stage C is known as
forward transduction or mechanical to electrical transduction stage and
stage D is known as Reverse transduction or electrical to mechanical
transduction stage.
 Briefly this is accomplished by the OHCs contracting upon upward
movement of the BM and expanding upon downward defection of BM.
9|Page

This expansion and contraction of the OHCs adds to the total

displacement of the BM. This added BM movement enhances amplitude
of the signal; thus the term cochlear amplifier used to describe this
function.

 Because the additional contraction and expansion of the OHCs occurs in

a restricted area of the BM, the peak of deflection is sharper in this area
and the frequency tuning is enhanced.
 Cochlear amplifier works at only low to mid intensity ranges and is
dependent on metabolic activity for influencing frequency selectivity.

2) Receptor potential nonlinearity:

The basilar membrane input/output function highly nonlinear at CF, but linear
above and below characteristic frequency (Rhode, 1971).

Level dependent nonlinearity

FIG: Level dependent non-linearity

Hudspeth & Corey (1977) explained displacement of stereociliary bundle Vs

hair cell membrane potential. He found compressive nonlinearity in both the
hyper and depolarization with respect to the receptor potential. He found that
the receptor potential got saturated earlier in the hyper-polarization stage than
depolarization stage which explains that there are more channels to be opened
than closed. This is the asymmetry in this non-linearity which makes our system
less loyal but in turn gives us the huge DR.
10 | P a g e

As a result we can see that for small hair bundle displacement the system is
good in recapitulation of input but when the hair bundle displacement becomes
huge it gives an output which gives us a squared off appearance similar to peak
rounding.

3) Two tone Suppression (TS)

Presence of one stimulus can reduce the responsiveness of the cochlea to other
stimulus. Originally called as “two tone inhibition” but “two tone suppression”
is preferred. Given by SACHS & KIANG, 1968.

 In a linear system, two tones with different frequencies do not interact.

The response to one tone has no effect on the response to another tone.
However, the cochlea is clearly a nonlinear system: In nonlinear systems,
responses to tone pairs often include interactions between the tones, such
as suppression and intermodulation distortion.

 Two-tone suppression consists of the reduction of the response to one

tone by the simultaneous presence of another.

 In non-linear system like cochlea, responses to 2 or more tones often

include interactions between the tones giving suppression and
intermodulation distortion. Two tone suppression consist of reduction of
the response to one tone by the simultaneous presence of another. Out of
the tones, one would be “probe tone” and second “suppressor tone”. So
it is the reduction of cochlear gain to the probe tone, due to the
simultaneous presence of the suppressor tone.
 Effect is greatest when the characteristic frequency & suppressor tones
are very close in frequency & decreases as their frequencies are further
apart. It is said that the two tone suppression at the level of auditory nerve
is also due to nonlinear processes in the cochlea (Glister, 1998), Robles
and Ruggero , 2001).

 Two tone Suppression is observed in the BM, auditory nerve and

electrical responses of individual inner and outer hair cells.
11 | P a g e

 Other intensity and frequency dependencies in two-tone suppression have

been described, but are beyond the current. One might also note in this
context that two-tone inhibition can affect auditory neuron tuning curves,
although these effects are different depending upon whether the inhibitory
tone is in close proximity to the CF or considerably below it
(KiangandMoxon, 1974; Javelletal.,1983;Patuzziand Sellick, 1984). For
example, the sharpness of tuning is hardly affected in the former case,
whereas broader tuning results from the latter. A considerable body of
evidence reveals that two-tone suppression is the result of nonlinear
processes in the cochlea (Geisler, 1998; Robles and Ruggero, 2001). For
example, two-tone suppression has been shown to occur for the vibratory
response of the basilar membrane (Rhode, 1977; Ruggero et al., 1992;
Cooper, 1996), cochlear microphonics and intracellular receptor potentials
(Pfeiffer and Molnar, 1970; Legouix et al., 1973; Dallos and Cheatham,
1974; Sellick and Russell, 1979) and otoacoustic emissions (Kemp and
Chum, 1980).
 Two tone suppression derives from the nonlinear responses of outer hair
cells as in the saturating input output function.
 When hair cells are stimulated with one stimulus, a superimposed stimulus
will drive stereocilia to greater displacement, ie, further to the flatter part of
input output curve. Because of the non-linearity, the response from the outer
hair to the two stimuli together, will be less than the sum of the responses to
the two stimuli considered separately.
 Therefore, Each stimulus will reduce the amplification of the travelling
wave to the other one & this results in the saturating non linearity of the
basilar membrane intensity functions.

4) Combination Tones/ Intermodulation Distortion

 Combination tones provide powerful evidence for nonlinearity.
 If the system is entirely linear, its output waveform will contain only the
frequency components as the input.
 If the system is nonlinear single tone will also produce harmonics.
 A pair of input tones will, in addition produce combination tones, that is,
tones whose frequencies depend on the frequencies of both the input
tones.
12 | P a g e

 When two tones are presented simultaneously, humans can hear

additional tones that are not present in the acoustic stimulus.
 If i/p-o/p function of the nonlinearity is symmetric around the line (even
order function) at input = 0 it produces an output with terms of the form n
(f1± f2) E.g., 3f2-3f1, 2f2-2f1, f2- f1… and even harmonics of the
primaries. (2f1, 4f1, etc….)
 If i/p-o/p function of the nonlinearity is antisymmetric around the line
(odd order function) it produces the primaries and combination tones of
the form nf1± mf2
 E.g., 4f1- 3f2, 4f2-3f1, 3f1- 2f2, 3f2- 2f1, 2f1-f2, 2f2-f1…..and odd
harmonics E.g., 3f1, 5f1…..

The combination tone is generated as a physical movement of the basilar

membrane at the site where the primaries overlap; and just as an external tone
sets up a travelling wave on the basilar membrane, the resulting oscillation
produces a travelling wave of its own, peaking at the place in the cochlea
corresponding to its frequency. There, the combination tone is transduced just
as though it were a tone that had been introduced from outside.

FIGURE : Diagram depicting the traveling-wave envelopes on the basilar

membrane for f1, f2, and the 2f1-f2. Note that the 2f1-f2 is lower in frequency
(and level) than the primary tones,but is generated where the two stimuli
overlap (shaded area) basal to the c 2f1-f2 characteristic frequency place.

5) Upward Spread of Masking

 The Upward Spread of Masking refers to the higher growth rate of
masking for maskers lower in frequency than the signal, compared to
maskers at the signal frequency (Wegel et al, 1924).
13 | P a g e

 The psychophysically measured upward spread of masking (USM) and

the neutrally measured two-tone suppression (2TS) are closely related
phenomena.
 They are alternative objective measures of the same OHC compressive
nonlinearity.
 USM involve the suppression of a basal (high frequency) probe due to the
simultaneous presentation of an apical (low frequency) suppressor. These
two views (USM versus 2TS) nicely complement each other, providing a
symbiotic view of cochlear nonlinearity.
 The travelling wave moves from the base to apex, the basal end of the
cochlea stimulated by low frequency sounds as well as high frequency
acoustic stimuli (Von Bekesy, 1947).
 When the low frequency sounds are relatively high intensity, the TW
action causes significant stimulation at the basal end of the cochlea,
which is likely responsible for Upward Spread of Masking (USM) effect
(the masking of stimulus at one frequency by a stimulus of a lower
frequency).
 The Upward Spread of Masking has most common clinical relevant in
hearing aid fittings. If an amplifying device provides too much low
frequency gain, the increased intensity of the low frequencies that is
provided by the hearing aid may mask the high frequencies, and overall
hearing performance may suffer.
 There are two parameters to describe the USM effect: υ and Im. The first
parameter, υ, is the slope of the masking as a function of masker level,
also known as the Growth of Masking (GOM). The parameter υ, thus
describes the “strength” of the masking. The second parameter, Im,
indicates that the masker intensity is at the threshold level or the lowest
level of the masker which causes masking of the probe.

Implications of Nonlinearities in Audition

The nonlinearity of aural responsiveness is associated closely with
1) Wide Dynamic Range
2) The Ears great sensitivity
3) Sharp frequency selectivity
4) Ability to generate otoacoustic emissions

1) Wide Dynamic Range

14 | P a g e

One of the most impressive features of the auditory system is its wide
dynamic range. In the mid frequency region between approximately 1 and 3
KHz,
listeners with normal hearing can detect pressure variations of approximately
0.00002 Pa and yet pressures one million times greater than this can be
reached before the sensation becomes universally painful or damaging. So
the dynamic range of human ear is 100-120 dB in the mid frequency region.
To accommodate such diverse stimuli, the ears responsiveness must be
nonlinear. This compression allows the extremely large range of levels in the
acoustic environment to be squeezed into a much smaller and
physiologically manageable range of responses.
The visual system also has large dynamic range but anyone going from a
light room to a dark room will testify vision can take several minutes to
return to maximum sensitivity after strong stimulation. In contrast the
auditory system recovers its full sensitivity within a fraction of a second.

.
2) The Ears great sensitivity:
Not only does the auditory system have an extraordinarily large dynamic
range but is acutely sensitive to small changes in level throughout that
range. The humans can detect small (approximately 1dB) changes over
this level over this entire dynamic range. While the dynamic range of the
normal cochlea is 100-120 dB, IHCs have a dynamic range of 50-60 dB
and Auditory Nerve about 20-40 dB. Cochlear compression solves the
IHC and AN dynamic problem.

3) Sharp frequency selectivity:

Another important feature of auditory system is the frequency selectivity.
For a sine wave of single frequency, the vibration has a sharp peak which
is confined to a narrow region of the basilar membrane.
 The enormous range of hearing (100-120 dB) and frequency
selectivity is accomplished via a form of compressive nonlinearity
that exists in the cochlea of the inner ear.
 The rate of changes in the cochlear responses decreases when the
sound pressure level becomes high. This process is known as
compressive nonlinearity.
15 | P a g e

 It constitutes a protective mechanism against too loud sounds. The

loss of nonlinear compression leads to loudness recruitement
(Steinberg and Gardner, 1937).

4) Ability to generate Otoacoustic Emissions (OAEs).

Clinical Application of OAEs:
- Neonatal screening
- Hearing screening-preschool and school aged children
- Diagnostic assessment
1. Auditory neuropathy
2. Monitoring cochlear function in ototoxicity
3. Noise and music induced hearing loss
4. Meniere,s disease
5. Malingering/ pseudohypacusis
6. Retro cochlear auditory dysfunction
7. Central auditory processing disorder