Transcript ppt presentation - Cochlear Implants

February
th
5
2002
Cochlear Implants for Reviewers
A presentation made to the ENT
section of the FDA
0
Cochlear Implants for
Reviewers
1: The Design of History
1
Early Efforts: The Tonotopic Theory
Based on the work of von Békésy, most
early cochlear implant researchers assumed that the only viable way to stimulate the cochlea was to place electrodes
along the basilar membrane and stimulate
the dendrites which were presumed to be
present: “place = frequency”
Dr. Fred Linthicum of House Ear Institute
subsequently demonstrated histologically that lack of
hair cells generally meant lack of dendrites, and it is
now universally assumed that the spiral ganglion cells
within the modiolus are the site of stimulation
2
Early Efforts : The Tonotopic Theory
Dr. House accepted the tonotopic theory, and, with
the help of some volunteer patients, implanted hardwired, 5 electrode systems to develop a usable
device
He found that his patients did not report great
differences between “place” (tonotopic)
stimulation and general stimulation (unpublished data)
This would not, could not be the case if it were
required to stimulate tonotopically
Other researchers reported similar results
This was very strange— inexplicable, in fact— but the
clinical result was undeniable
3
Single Electrode Implants: Overview
As a result of these clinical results, the first single
electrode cochlear implants— indeed, the first
practical cochlear implants of any sort— were
developed
Incoming sounds were amplitude modulated on a full-time,
16-24 kHz signal and injected into the cochlea
The field developed is apparently present globally,
stimulating all spiral ganglion cells and nerves
simultaneously
These are “analog presentation” cochlear implants
4
Single Electrode Implants: Overview
From the beginning, it was assumed— without clear evidence and
for reasons that we will explore— that multiple electrode (“multiple
channel”) cochlear implants would work better
“Until… multichannel prostheses become a reality, one must consider
the question of whether or not it is reasonable to continue implanting
single-channel prostheses.”
“Above all, a single channel auditory input will not provide a speech
input that either sounds speech-like, or is understandable.”
– Bilger RC, et al. Implanted auditory prosthesis an evaluation of subjects
presently fitted with cochlear implants. Trans Am Acad Ophthalmol
Otolaryngol 1977 Jul-Aug;84 (4 Pt 1)ORL-677-82.
The Bilger report on a handful of early single electrode patients
inferred these criticisms in 1977, before any multiple electrode devices
were even available for comparison, and no data provided support
such conclusions; they appear to be based entirely in theory
5
Early Efforts : Multiple electrode devices
In such a climate (as indicated by the Bilger
quotes) it is not surprising that other researchers
started by building multiple electrode devices;
they never tried any alternatives
With early designs, it was impossible to stimulate the
cochlea at multiple sites simultaneously because of
unwanted signal interactions
Because of the capacitive nature of the cochlea,
stimulation had to be followed by “radio silence” to
allow charges to drain away
These devices used pulsatile stimulation, shifting
rapidly between frequencies (sites of stimulation)
6
Early Efforts : Multiple electrode devices
Multiple electrode implants:
Incoming sounds are separated into different frequency bands
Each frequency band is assigned to a different electrode pair
When stimulating for the higher frequencies, the frequency of
the stimulating signal has not been thought to be important
(i.e. 1200-1500 pps is common)
Stimulation is pulsatile and discontinuous (staccato)
These are “pulsatile presentation” cochlear implants
7
Two Basic Kinds
As a result of this history, there are
two basic kinds of cochlear implants
multiple, long electrode
(pulsatile presentation)
single, short electrode
(analog presentation)
8
Cochlear Implants for
Reviewers
2: Design, Function, Surgical Placement
9
Single Electrode Implants: Design
Many different kinds of “analog presentation”
implants can be imagined, but the specific
design of the current AllHear devices is rather
simple
The internal receiver consists of few parts:
10
Single Electrode Implants: Design
The external processor, as yet, is likewise
very simple:
Electronics
package
Volume
control
Battery
Microphone
Coil
Magnetic
insert
Coil
11
Single Electrode Implants: Design
The external processor amplifies sounds and combines
them with a 16-24 kHz carrier (amplitude modulation)
The AC output is sent through the external coil, which
develops an alternating magnetic field
This field interacts with the internal coil, inducing a
corresponding AC current which is sent through the
electrodes
The present design processor uses “clipping” to control
overloud signals
12
Single Electrode Implants: Design
13
Single Electrode Placement
Insertion of no
more than 6mm
preserves cochlear
structures and
residual hearing…
14
Multiple Electrode Placement
“Deep insertion”
of 20mm or more
is known to
destroy cochlear
structures and
tends also to
damage residual
hearing…
15
Multiple (long) vs. Single (short)
The contrast
between the two
provides for
greater safety
during the surgical
procedure
16
Cochlear Implants for
Reviewers
3: The Perception of Percept
17
Single Electrode: the perception of percept
All presently manufactured cochlear implants
approved for general use in the US, Europe and
Japan use multiple electrodes
If, as Dr. House found, there is little difference
between patient percept whether using one or
several electrodes internally, where external
processor electronics are held constant..
Why is this the case?
Why does “everyone know” that multiple electrodes are
better?
The answer is the result of a cascade…
18
Single Electrode: the perception of percept
Early 3M/House processors were not designed to
provide good speech reception, they were designed
only to provide access to environmental sounds
This, because at the time no one thought that these
new devices– cochlear implants– would actually
provide patients with good enough sound so they
would understand speech
Thus incoming sounds were deliberately restricted in
frequency bandwidth, reflecting safety concerns, and
these low expectations
This became a self-fulfilling prophecy, as a circuit
diagram of the original device shows…
19
Single Electrode: the perception of percept
Fretz RJ, Fravel RP. Design and function: a physical and electrical description of the
3M House cochlear implant system. Ear Hear 1985 May-Jun;6(3):14S-19S
20
Single Electrode: the perception of percept
Restricting frequencies to below 2700 Hz— eliminating
some 35% of the frequencies important to
understanding English— means that consonant sounds
would be more difficult to hear
Indeed, these processors had several other inherent,
performance-limiting problems:
They did not use modern compression, but rather “clipping”,
which tends to introduce increasing distortion at full volume
They had noisy circuitry, degrading the S/N even in quiet
conditions
…And they had other, “non-electrode” problems
21
Single Electrode: the perception of percept
If we were to compare two hearing aids, one of which
offers the full range of frequencies important to
understanding speech, and the other of which deliberately
restricted frequencies (excluding some 35% of the
information in speech)…
…one of which offers compression and the other of
which does not…
…one of which offers a relatively clean signal and the
other of which has considerable added noise…
With which one would we expect patients to do better?
22
Single Electrode: the perception of percept
The 3M/House and 3M/Vienna devices were termed
“single channel”, and this was assumed to be their pivotal
“defining” characteristic, even though (for example) the
Vienna device was extra-cochlear!
The limits of the processor were not considered, and no
data exist which offer information about how changes to
single electrode, analog presentation processors affect
patient percept
Does it make sense to assert, without one scintilla of
supporting evidence, that the perceptual limits
experienced by patients were (solely, or even primarily)
due to the use of one electrode?
23
Single Electrode: the perception of percept
To gain a proper perspective on this subject, it is
crucial to realize that every theoretical objection to
these devices, on analysis, centers around the
assumption that good frequency information cannot
be provided when using a single electrode
Sound is pitch, timing, loudness; frequency is the pivot
Thus if it can be shown that such devices can
provide good frequency information…
then previous assumptions are shown to be wrong…
…and any prejudice against these devices centering
around the electrode must be re-examined and perhaps
discarded
24
Single Electrode: the perception of percept
The issue of quality of percept is obviously
central, but putting it aside for the moment,
let’s examine some other facets of the
comparison…
25
Cochlear Implants for
Reviewers
4: Tono topics
26
Considering the T-Theory
For the sake of this discussion, let’s assume
that the tonotopic theory can be stated as:
We must stimulate the cochlea according to the
expected “place” of the frequency-of-interest, in
order to provide a percept of that frequency
Please note the use of the word “must”,
which clearly informed Dr. Bilger’s idea about
this theory:
“…single channel auditory input will not provide…
speech… that is understandable…”
27
“Channels”: a slight digression
“Above all, a single channel auditory input will not provide a speech
input that either sounds speech-like, or is understandable.”
Simply because we have created a name for something, does
that thing exist? (“multiple channel”, “single channel”) What is
a channel?
It would seem that either “a channel” is a single frequency (or
narrow band noise), or it has no meaning whatsoever
For multiple electrode CIs it may make sense to refer to
“channels”, but it makes no sense when referring to analog
presentation CIs
We believe we can demonstrate that “single channel” as a
term applied to single electrode devices has no scientific
content
28
Considering the T-Theory
The tonotopic theory appears to work, based
on the success of present devices, and
studies showing that the place of stimulation
within the cochlea is associated with different
pitch percepts
But is there evidence for the possibility of
successful non-tonotopic stimulation?
And if so, how would one fit these two
(different?) modes of stimulation into a “unified
theory” of cochlear stimulation?
29
Considering the T-Theory
In fact there is good evidence for the success of nontonotopic stimulation
For example, in Appendix G of The Audiologist's
Handbook for the Mini System 22 (Cochlear Corporation,
April 1993) the following quote is found:
“The [Cochlear Corporation's Mini 22] cochlear implant can
produce two types of stimulation, bipolar and common ground
(CG)... Typically, the lowest thresholds are obtained in CG
stimulation because of the wider current spread.”
“Common ground” means currents spread globally— nontonotopically— within the cochlea
30
Considering the T-Theory
“Wider current spread” engulfs more nerves—
but studies done comparing CG to bipolar for this
device do not appear to show significant
differences in percept
“…results indicate that… restriction of the size of
the neural population activated by individual
channels of the prosthetic is not necessarily
advantageous.”
– Pfingst BE, Zwolan TA, Holloway LA. Effects of stimulus
configuration on psychophysical operating levels and on
speech recognition with cochlear implants. Hear Res 1997
Oct;112(1-2):247-60
31
Considering the T-Theory
“Bipolar” stimulation is stimulation of closely paired electrodes;
“monopolar” stimulation is stimulation of an active electrode
coupled to an extra-cochlear ground
Clearly, monopolar stimulation would lead to wider current
spread, and should reduce frequency discrimination because it
“breaks” tonotopicity…
“An assumption that bipolar stimulation should provide better speech
understanding… has not been confirmed in this study. We found… no
significant difference… on any tests when subjects were compared using
monopolar or bipolar stimulation modes and the same coding
strategy…”
– Battmer RD, Martens U, Gnadeberg D, Hautle K, Lenarz T. Comparison
study of patients using either the Nucleus Minisystem-22 in bipolar mode
or the Nucleus 20 + 2 in monopolar mode. Ann Otol Rhinol Laryngol
Suppl 1995 Sep;166:349-51
32
Considering the T-Theory
As well, no long electrode, pulsatile presentation
implant currently in production reaches more than 25
mm into the cochlea, whereas the basilar membrane
is generally quoted as being 35 mm long
That is, no pulsatile presentation cochlear implant
stimulates the apex of the cochlea, where low
frequencies (1500 Hz & lower) are apparently
resolved
Yet patients using these devices apparently hear and
resolve these low frequencies…
33
Considering the T-Theory
Some proponents of the tonotopic theory will tell
you that such patients “hear” unstimulated low
frequencies because such frequencies are at or
below the firing rate of the nerves (~500 cps), or
that some other mechanism (volley theory)
allows for that percept
But does it make sense to assert that the cochlea
has one mechanism to hear low frequencies, and
another to hear high frequencies?
34
Considering the T-Theory
Occam’s razor: Rather than assuming, simply
to safeguard a tenuously-based theory, that
there are two mechanisms, it would make
more sense to assume that there is only one
mechanism being exploited in several ways…
But if there is only one mechanism, what
might it be?
Consider the work of Kiang…
35
The Studies of Kiang
Kiang recorded
from an electrode
in a single auditory
nerve axon in intact
cat cochleae
Over a period of
years recordings
were made from
hundreds of VIII
nerve axons
36
The Studies of Kiang
Kiang then presented the cat’s ear with a
stimulus at a certain decibel level (loudness),
where the tone (frequency) was increased
continuously
While this steadily rising tone at a given
decibel level was being presented to the cat’s
ear, Kiang recorded the frequency at which
the single nerve fiber in which he had his
probe began to fire, and the frequency at
which it stopped firing
37
The Studies of Kiang
Then he repeated the “rising tone” stimulus at a
lower decibel level, with his probe in the same nerve
fiber
Eventually, at a low enough decibel level, he found
that each nerve would fire at only one very specific
frequency, which he called “the characteristic
frequency”
At a given decibel level then, Kiang found out the
range of frequencies to which a specific nerve fiber
would react, and he produced charts of those
reactions…
38
The studies of Kiang
Recording with stimulus at 60 dB
A specific nerve fiber reacts from 180 to 900 Hz at 60 dB
39
The studies of Kiang
Recording with stimulus at 40 dB
The same nerve fiber reacts from 330 to 500 Hz at 40 dB
40
The studies of Kiang
Recording with stimulus at 20 dB
The same nerve fiber reacts only to 490 Hz at 20 dB
41
Kiang’s conclusions
If the volume of the stimulating sound is high
enough, a nerve fiber will fire across a broad range of
frequencies
When the volume of stimulation is reduced, the nerve
fires across a narrower range of frequencies
At the lowest volumes where the nerve will stimulate,
only one frequency causes that nerve— and few if
any others— to fire
Again, Kiang called this its “characteristic frequency”
42
Speculation…
Obviously Kiang was working with sound stimulus, not
electrical stimulus
Even still, surely his work provides a useful means of
thinking about electrical stimulation, because it seems
sensible to assume that some or much of the preference
for frequency resides within the nerve itself, and the nerve
is an electrical organ…
And if the nerve has “preference”, might it not be that
intense pulsatile electric fields can overwhelm the nerve’s
characteristic frequency, whereas less intense, analog
fields tend to allow that preference?
If so, then place stimulus must be intense stimulus, but
more benign global stimulus may be analog in nature
43
Speculation…
That is, tonotopic, pulsatile stimulation works
by overwhelming the nerve’s tendency to
respond only at its characteristic frequency…
It can be stimulated, if we hit it hard enough
…and analog stimulation works precisely by
exploiting characteristic frequency …
It will respond to its characteristic frequency, if
present
44
Cochlear Implants for
Reviewers
5: “Doctor, is there a pulse?”
“Digital electronics deals with 1's and 0's, the language of logic,”
explains Tilde, “whereas analog electronics deals with waves, the
language of sound.”
Smithsonian Magazine, February 2000, “Redefining Robots” p 96
45
Using pulses: a slight digression
The pulsatile stimulation used by CIs is locally very intense
The purpose is to cause all nerves local to the stimulus to “fire”
simultaneously; to “entrain” them; to stimulate a “volley”
Consider: What this means is that the stimulation involved must
necessarily result in either a single tone, or narrow band noise
Remember: it is apparently not the frequency of stimulating signal
that provides the result… It is the local intensity
The number of electrodes therefore equals the number of
frequencies or “channels” (CII possible exception)
Multiple electrode devices therefore offer a “set” of discrete
frequencies, not a continuous range of frequencies
46
Using pulses: a slight digression
So, are nerves “digital”? Are pulses the best way
to interact with them?
In fact, recent research is beginning to uncover
the much more subtle, entirely more ubiquitous
and rather complex analog nature of nerves
Why should we believe that the inner ear, which
has never naturally produced nor experienced
anything like a sharp rise-time, square wave
pulse, would offer a better response to a
pulsatile vs. an analog stimulus?
47
Using pulses: a slight digression
Beyond the fact that pulses were required to get multiple
sites stimulated in rapid and sequential fashion, there
were historical reasons for using pulses
Early (and continuing) research done on nerve stimulation
required large artifacts in order to “discover” nerve
responses in the electrically noisy biological environment
Recordings were made of the response to hundreds or
thousands of stimuli, the results were time normalized and
averaged, and a waveform resulted
Sharp rise times— which characterize pulses— were
therefore the stimulus of choice, because they invoke
large artifacts which are easy to see, “on average”…
48
Using pulses: a slight digression
But this is the scientific equivalent of looking for one’s
keys under the streetlight
Certainly the approach yielded results, but it also
encouraged researchers to believe that the only nerve
responses worthy of research were large affect
responses— the ones which were easiest to see
Yet such responses are arguably completely artificial, and
might only be produced by using pulses to invoke them,
meaning that at best they offer us a picture of a nerve
under considerable stress…
The generally unchallenged idea grew up that nerves are
digital, they “fire”, they are either on or off
49
Using pulses: a slight digression
“Jerome Lettvin… MIT professor of both physiology and
engineering… says, ‘We are not digital machines, we are
analog devices. All our neurons are analog devices, and
anyone who tells you otherwise is talking nonsense.’ ”
– Smithsonian Magazine, February 2000, “Redefining Robots” p 96
Digital technology throws away direct information to increase
derived accuracy; analog technology relies on direct
information, often at the expense of derived accuracy
Digital systems measure; analog systems compare
Sensory systems appear generally to prefer more information
(and comparison) over greater accuracy (and measurement)
50
Cochlear Implants for
Reviewers
6: Single vs. Multi
Less is more, more or less?
51
Multi vs. Single: Established Facts
A single electrode CI design has a number of
advantages over multiple electrode designs,
in several important areas
•Preservation of hearing
•Reduced expense
•Ease of setting
•Higher bandwidth
•Cleaner signal
52
Cochlear Implants for
Reviewers
6a: Single vs. Multi
53
Multi Disadvantages: Setting Devices
Properly setting multiple electrode implants is very difficult and
expensive
16-22 active electrodes, 4 key parameters each, some or all
electrodes can be fired, ganged or single electrodes, several
different “strategies”; all interactive and inter-related, one to
another, and therefore highly complex
Initial “mapping” generally takes several sessions, consuming
between 5 and 20 hours; and it may take longer…
“Refining the programming of a cochlear implant can be a tedious
process that can take a year or more of hard work before patients can
hear clearly.”
– Chris Bertrand, CEO of MED-EL North America
54
Multi Disadvantages: End-User Expense
Multiple electrode implants are expensive
The standard quoted price for multiple
electrode implants is $50,000 to $75,000
(implant cost: >$20,000)
The average per capita yearly income in the
US is $36,000
The average per capita yearly income in
China is $1,500
55
Multi Disadvantages: Single channel?
Because of “unpredictable fluctuations in
loudness” if multiple signals are injected
simultaneously into the cochlea, then for the
market leader CI, only one electrode pair (or
a set of ganged electrodes) is fired at any
one moment
Thus from the point of view of time, the
device is “single channel”
56
Multi Disadvantages: “Bandwidth”
The concept of bandwidth has to do with the amount of
information one can provide across a certain link or pipe in
a given unit of time
For CIs, bandwidth would have to do with…
the abstract ability of the damaged hearing system to resolve pitch,
timing and loudness from the best possible information, and…
the abstract ability of the CI to provide pitch, timing and loudness
cues
To the degree possible, one must try to separate the limits
of the hearing system from the limits of the mechanism
Because we are always measuring the latter via the means
provided by the former, we as yet have poor information
about the true potential of the damaged hearing system…
57
Multi Disadvantages: Limited Bandwidth
We can, however, infer some things about the
bandwidth of multiple electrode CIs
Only a portion of the frequency spectrum is
provided at any one moment
Time is also underutilized, as stimulation is followed
by silence (staccato)
Depending on how the device is programmed, it
may only be able to provide 5% to 60% of the
information available in the full ambient signal
58
Multi Disadvantages: Limited Bandwidth
As mentioned, when we try to discover the bandwidth
limits of the damaged hearing system, we should
recognize that we may be measuring the mechanism, not
the system
An example is provided by studies which show that
between four and ten electrodes are all that is required for
speech reception
Such studies depend entirely on the specific forms of
pulsatile processing which were used, so it seems
unwarranted to assume that we have learned very much
about the underlying system
The means used are also language-specific, and this again
indicates they are not fundamental enough…
59
Multi Disadvantages: Limited Bandwidth
“Speech processing” (SP) overcomes bandwidth limits to some
degree, generally by sacrificing full access to a variety of
sounds
SP has inherent bias with regard to language
This is not to say that SP is a poor choice; it may be
necessary, particularly for those with the most damaged
hearing systems, even if the means of stimulus provides better
bandwidth
SP is highly likely to be more effective with greater device
bandwidth
But the “philosophy” of multiple electrode CI design is
fundamentally different from the philosophy of hearing
aids, where clarity at volume is the primary goal
60
Multi Disadvantages: Noise Is Inherent
Because multiple electrode designs stimulate the
cochlea using pulses, they cannot faithfully
recapitulate analog waveforms within the cochlea
“Aliasing” must result, essentially introducing
distortion or noise into the signal provided
It appears that the only way to overcome such limits
is to move toward fully analog signals
Higher rate pulsatile systems are a step in this
direction, and we would expect them to show better
results because of increased bandwidth and clarity
61
Multi Disadvantages: Noise Is Inherent
“Because multiple electrode designs stimulate the
cochlea using pulses, they cannot faithfully
recapitulate analog waveforms”
62
Multi Disadvantages: Cochlear Destruction
Numerous studies have shown that the
insertion of a multiple electrode– long
electrode– CI destroys irreplaceable cochlear
structures, including the basilar membrane,
hair cells, and spiral ganglion cells
“There is general agreement that [such]
damage… may lead to severe neural
degeneration.”
Kennedy DW: Multichannel intracochlear electrodes: mechanism of insertion
trauma. Laryngoscope. 1987 Jan;97(1):42-9.
63
Multi Disadvantages: Cochlear Destruction
Other studies have shown that the intense stimulation used in multiple
electrode devices, in itself, might be destructive
“Earlier studies established damage thresholds for both acute
(…70 µC/cm2) and chronic (…15-20 µC/cm2) stimulation…”
Duckert LG, Miller JM: Mechanisms of electrically induced damage after cochlear implantation. Ann Otol Rhinol Laryngol. 1986
Mar-Apr;95(2 Pt 1):185-9.
“…the charge density for the analog sinusoidal output at 1000 Hz
is 379 µC/cm2… As the stimulus frequency drops, the charge
density increases until it reaches 667 µC/cm2…”
Advanced Bionics IDE submission
667 µC/cm2 is nearly 10 times the acute damage threshold reported
by Duckert, and approximately twice the level at which electrode
gassing has been reported to occur (by Brummer)
64
Multi Disadvantages: Cochlear Destruction
This is another consequence of using these very intense
pulses, and further evidence of the fact that they are
biologically unknown…
However, since we do not have significant numbers of
reports of long-term degradation of percept (which we
might expect in the case of destructive stimulus), we have
to conclude that the cochlea is more tolerant than we had
thought, or that the damage being caused does not
degrade percept, as might be the case depending on the
mechanism of stimulation and the site of the damage
Clearly the final word is not yet in…
65
Cochlear Implants for
Reviewers
6b: Single vs. Multi
66
Single Advantages: Setting Devices
Properly setting single electrode analog implants is very
easy, and inexpensive
“Setting” is essentially the same as adjusting a hearing aid
Thus the expertise required is available in every city and
town, wherever there is a HA professional
Everything needed can be put into a standard desktop
computer (NOAH), or a handheld (Palm Pilot), and can be
made prescriptive (NAL-NL1, etc.)
This has profound implications for the distribution of
cochlear implants, and the shape of the market…
67
Single Advantages: End-User Expense
The standard quoted price for single electrode
implants is $15,000 (implant cost: $6,0008,000)
We believe it would be possible to develop a
low-cost ($1,500) device
No other cochlear implant design type could
possibly penetrate the world market…
68
Single Advantages: Full Bandwidth
While the limits of the damaged hearing system remain
unknown, there is evidence that the analog
presentation approach holds promise
Single electrode analog implants apparently provide
access to many or all frequencies (more information momentarily)
Single electrode analog presentation cochlear implants
can provide
The full frequency spectrum
Full-time, with
Pure analog presentation
69
Single Advantages: Noise is Avoidable
Single electrode analog implants can
apparently provide clean signals
In advanced hearing aid technology, digital
representations are converted to analog
sound for presentation via a speaker (“Digital” is
good, when restricted to the inside of an amplifier)
Advanced single electrode analog implants
will work in exactly the same way, except
they will provide output to a coil
70
Single Advantages: Cochlear Preservation
Insertion of short electrodes generally
preserves cochlear structures
“Thus, there is no evidence that the short 6-mm
electrode used in the 3M/House implant will
necessarily destroy residual hearing that might be
present in a child receiving this implant.”
Dye et al: Measurable Residual Hearing Following Cochlear Implantation. In Meyers et al:
Advances in Otololaryngology Vol 4; St. Louis Mosby Year Book: 1990 61-79
Statistics are as yet unavailable, but it may be that the rate
of iatrogenic hearing loss from short electrodes is roughly
equivalent to stapes surgery, which is generally <1%
Stimulation levels are <1 µC/cm2
primarily because stimulation never falls below 16 kHz
71
The Matter of Proof
Clearly, single electrode, short electrode, analog
presentation cochlear implants have a number of
significant medical and strategic commercial
advantages
The facts presented prove greater safety, and they
demonstrate the considerable potential of this
technology to be improved
But is there strong evidence of efficacy?
Can patients using these devices gain a good
percept?
72
Cochlear Implants for
Reviewers
7: Proof of Life
73
The Matter of Proof
Again, please remember: it is crucial to realize that
every theoretical objection to these devices, on
analysis, centers around the assumption that good
frequency information cannot be provided when using
a single electrode
“Above all, a single channel auditory input will not provide a speech
input that either sounds speech-like, or is understandable.”
Please also note that we are discussing
evidences of the potential of a technology, not
proof of the efficacy of a device which is only
one of several possible implementations of that
technology…
74
The Matter of Proof: “MF”
Age 19, implanted age 15 mo. with 3M/House,
upgraded to AllHear processor in 1993
Favorite HS class German, plays the saxophone
Daily telephone use
Evaluated at University of Iowa by Dr. Bruce Gantz:
Test results
34.7% HINT sentences
38.0% CNC words
58.7% CNC phonemes
These scores are impossible to a patient who is
getting only one or a limited number of frequencies
75
The Matter of Proof: “MF”
Dr. Gantz’s team reported that:
“[MF] is an extraordinary young man who has
developed excellent speech, language and reading
skills such that he does better on standardized
testing than almost all of his peers who have had
hearing all their lives…”
MF achieved these results using a processor
without modern compression, using clipping
What might he do with a better processor?
76
The Matter of Proof: “JF”
Age 20, implanted age 3 with 3M/House, upgraded to
AllHear processor in 1993
Daily telephone use
Plays guitar
Evaluated by Dr. Gantz’s team: Test results
40.8% HINT sentences
22.0% CNC words
34.0% CNC phonemes
These scores are impossible to a patient who is getting
only one or a limited number of frequencies
77
The Matter of Proof: “JF”
Dr. Gantz’s team reported that:
“…[JF] has developed excellent speech, language
and reading skills such that he does very well on
standardized testing, when compared with his
peers who have had hearing all their lives…”
JF achieved these results using the presentdesign OTH
What might he do with a better processor?
78
The Matter of Proof: “JR”
Age 22, implanted age 5 with 3M/House,
subsequently upgraded to AllHear processor
Schooled without an interpreter
Daily telephone use
Evaluated by Dr. Gantz’s team: Test results
59.3% HINT sentences
18.0% CNC words
43.3% CNC phonemes
These scores are impossible to a patient who is
getting only one or a limited number of frequencies
79
The Matter of Proof: “JR”
Dr. Gantz’s team reported that:
“…[JR] has developed excellent speech, language
and reading skills such that he does very well on
standardized testing, when compared with his
peers who have had hearing all their lives…”
JR achieved these results using the OTH
What might he do with a better processor?
80
The Matter of Proof: “TL”
Age 22, implanted age 3 with 3M/House, upgraded to
AllHear processor in 1993
Honors student at her university (now graduated)
Attended classes with no interpreter
Daily telephone use
“ …(TL) is a superior user… Her performance on a modified version of
the Glendonald Auditory Screening procedure [was equaled by only]
1.5% of the total 126 subjects…”
Chute PM, Hellman SA, Parisier, SC, Selesnick SH: A matched-pairs comparison of single and
multichannel cochlear implants in children;Laryngoscope 100 January 1990, p 25-28
This was in 1990. She attained this result using a 3M
processor.
81
The Matter of Proof: “TL”
[TL] is an extraordinary user, without question. So were
the two other children who equaled her performance in
the Chute study. We have three extraordinary users using
two different kinds of implants
However, to assume that [TL] was so extraordinary that,
using a cochlear implant which provided only one
frequency, she was able to equal the performance of the
very best users from among this relatively large group, is
completely beyond belief
Her score would surely have been impossible to someone
getting very limited information from their CI
82
The Matter of Proof: Many patients
“We evaluated the ability of profoundly deaf
children using the 3M/House single-channel
cochlear implant to understand speech without
the aid of speechreading… Fifty-two percent
demonstrated some open-set performance on
word identification…”
Berliner KI, House WF, Tonokawa LL: Open-set speech recognition by children with a single
channel cochlear implant; Transactions of the American Oto Society, 1989
Again, these children were using processors with known
and significant deficiencies…
83
The Matter of Proof
Thus it is clear that a single electrode can provide
access to many frequencies
Indeed, to generalize, one may say that we have an
adequate demonstration that a ubiquitous AC current in
the cochlea can apparently provide a percept of sound
This demonstration completely negates the objection
lodged against single electrode devices so many years
ago, and unquestioningly continued through the present
That is, we have shown that “single electrode” is far
different from “single channel” (i.e. 1 electrode = 1 frequency)
84
The Matter of Proof
In sum, patients using single electrode devices,
indeed when using external processors with
significant internal noise & distortion, and
without benefit of modern compression, have
demonstrated strong abilities; some are in the
first rank of cochlear implant users
What might they do with superior technology?
85
Cochlear Implants for
Reviewers
8: Summa
86
Multiple Electrode Implants…
…are very sophisticated devices, known to work
relatively well for patients
They are also:
Complex (internal electronics)
Expensive ($20,000-$30,000)
Destructive (long electrodes destroy the inside of the cochlea)
Difficult and expensive to set (requires computer)
Power hungry (new or recharged batteries every day)
87
Single Electrode Implants…
…are relatively simple devices, and have been
shown to work relatively well for some
patients, although statistical proofs are still
lacking
As well, they are:
Simple (no internal electronics)
Less expensive ($6,000-$8,000)
Safe (short electrodes preserve the cochlea)
Easy to set (requires screwdriver or Palm Pilot)
Power conservative (batteries last 10-14 days)
88
89