Cochlear Implants: My Perspective

By William F. House, D.D.S., M.D. (Edited by David House)

When science, art, literature, and philosophy are simply the manifestation of personality they are on a level where glorious and dazzling achievements are possible, which can make a man's name live for thousands of years. But above this level, far above, separated by an abyss, is the level where the highest things are achieved. These things are essentially anonymous

~Simone Weil
"Human Personality"

Chapter IV, Auditory Success

I CAN REMEMBER WHEN the goal of runners was to achieve the four-minute mile. For a good many years it was widely regarded as impossible, but finally Roger Bannister broke this "speed limit" in May of 1954.

Today one almost never hears reference to this measure of distance and speed, and a number of runners have exceeded this benchmark. This been accomplished first by setting certain goals, and second by studying and instituting all the factors which apparently contribute to attaining the goal: physical and mental training systems, nutrition programs, and so on. In fact Bannister not only broke through this barrier, but he established the pattern for others who followed, because he is said to have accomplished his task through scientific training methods and the study of the mechanics of running. (As an interesting aside, Bannister later became a neurologist.)

At this point in the history of cochlear implants, I think we are now ready to set the ultimate goal for implant patients as:

The common or universal attainment among implant patients of fluent open-set discrimination and successful functioning in normal auditory communication situations.

In essence, we want as many of our implant patients as possible to hear as normally as possible. This would be the four-minute mile or moon-shot of cochlear implants. Because of its nature and broad reach, this will be an anonymous victory, requiring the efforts of many people over a long period of time, and more in that respect like the moon landing than the four-minute mile.

For any given patient, once the implant has provided the best sound possible to a given patient, the next step is to achieve open-set recognition. [60] I refer to this as auditory success. Fortunately, we are seeing an increasing number of implant patients attain auditory success: about 80% of implant patients exhibit some degree of open-set recognition.

Clearly, we must continuously improve our record so that more and more of our patients achieve this aim, gradually winning our way to universal, or nearly universal, auditory success.

In attaining any such goal we must first identify the factors that probably contribute to the goal, and second to learn to enhance those factors. Absent better studies of course, we cannot as yet say definitively or precisely which factors contribute most to auditory success. However, some factors clearly have an effect.

The contribution of hardware

Above, I covered in some detail the general question of what the hardware provides, that being access to the critical parameters of sound.

As well, there was considerable discussion regarding electrode design, with the summation being that the design of electrodes according to the tonotopic theory is unreasonable and dangerous to residual hearing. Beyond this, the evidence is that either short or long electrodes can deliver all of the critical parameters of sound: so the subject of electrode design is not further pertinent to a discussion of what the hardware can offer the patient.

However, no mention was made of processing schemes, beyond a discussion of the early work which Jack Urban and I did. This area, concerning the manipulation of the sound signal before delivery to the internal implant, is, in my opinion, where the focus of our efforts regarding implant hardware must be.

There is considerable reason to believe the audiogram can be improved, or the AI scores can be very positively affected if you will, by the external sound processing scheme used in the implant system in question.

Consider an interesting finding in the Cohen et al study: [61]

Performance with implant 2 [the Nucleus 22] improved after the processor was changed. The mean composite index of the 24 patients who received MSP processors increased from 56 after they had used the WSP processor for 24 months to 86 (P>0.001) after they had used the MSP processor for 3 months. When the MSP processor was used, the mean score for open-set word recognition was 25 percent (range, 6 to 58 percent), and the score for sentence recognition was 58 percent (range, 9 to 97 percent).

To reiterate, after two years of using the WSP processor the composite score was 56; after three months of using the MSP processor, this increased to 86. In other words, without any change of internal hardware, the change of the external sound processing scheme had a rather significant effect.

This significant improvement apparently occurred because the implant user was provided with better high frequency thresholds. We are not given the pure tone measurement or AI information which would have demonstrated this beyond question, however. (Again, we must recognize the value of these metrics, and include them in all relevant reports in the field.)

As indicated previously, the frequencies from 2 to 4 KHz are of particular importance for proper discrimination of speech. For example, Skinner et al [62] demonstrated improved consonant recognition by emphasizing the high frequencies when the patients' Nucleus processors were changed from the Mini Speech Processor (MSP) to the Multi-Peak (M-Peak) processor.

Given that many such experiences have demonstrated the central value of the sound processing scheme, it bears mention that the 3M/House processor had, in essence, no processing scheme whatsoever. Sounds were simply amplified, without compression, a 16 KHz carrier was added, and that signal was induced in the internal coil. Further, the 3M/House processor had a frequency limiter in it so that any sounds above 3 KHz were simply thrown away.

The remarkable thing is that any patients using such equipment-- that is, with a primitive processor, albeit, as it turns out, with an advanced electrode — were able to have open-set recognition: but many were.

Again, this demonstrates the degree to which non-hardware factors can influence a patient's ability to become a successful implant user, and, again, it shows that comparisons between this older generation processor and newer processors — with schemes which provide for volume compression, feature extraction and the like — are invalid when applied to the question of electrodes.

In sum however, I am convinced that the most fertile field for the improvement of cochlear implants lies in a thorough exploration of the question of processing schemes. How can we best emphasize the relevant features of speech, using the hearing aid processor technology we now have available, in such a way that what is offered to the user gives optimum access to the sounds of speech?

The other very important implication of current information is that we can, for the foreseeable future, virtually ignore the electrode as regards design issues. A short, safe electrode of very simple design and efficient power utilization clearly offers us everything we need as a platform for almost any implant system. Therefore the area of greatest mystery, challenge, and potential for progress in the improvement of implant hardware will be here, in the design of smaller external processors and the discovery of better sound processing schemes.

The great advantage of this approach is that the electrode, once demonstrated safe, can serve as the reusable engine of progress in the FDA approval process. New processors will generally be seen as non-significant risks, and thus can gain approval, on the basis that they can provide comparable AI scores, much more rapidly.

Should this situation come to pass, a thousand flowers could indeed bloom.

The contribution of the patient

Clearly, however, if there is one thing which our experience with implants has proven to us by this time it should be that the most significant factor in the equation leading to auditory success is the patient. The fact, for example, that so many patients, who apparently have such a poor choice of sound cues are nevertheless able to demonstrate a high degree of open set recognition should be a testament to the amazing capacity of human beings. I stand in awe of some of these people, who apparently have an intelligence which we have not yet described, much less measured.

In a way, of course, this defines our challenge, for, in order to insure that we have universal or nearly universal auditory success, we must begin to describe and to discover ways to enhance these native abilities. We can do this in several ways.

For example, we should make efforts to study, consult with, and understand those who show extraordinary capacity. While some of what they have accomplished is likely due to skills and abilities they were born with, it may also be true that they have discovered strategies which help them better use the implant.

Further, we can search the literature in related fields, and consult with recognized experts regarding the problems we face.

Finally, we can look at the data we already have, and develop more data as well, which draws out the comparison and distinction between what the implant is providing (via careful recording and reporting of the AI scores of our patients), and what the patients are able to do with this information (via the speech scores which the patients in various programs achieve). Thus the baseline of comparison between implants, in any effort to identify ways to enhance patient skills, becomes the AI score, and the yardstick of achievement, placed on that baseline, is the patient's speech score.

If, on this basis, more successful programs and better processors emerge, we can examine them in order to discover the elements of their success, and blend these elements into existing programs and processors for the further benefit of our patients.

Patient parameters

Certain indicators of success have emerged, and I would like to mention them here without a lot of supporting detail, because I will be offering that information in later papers. Thus, briefly, we know that:

• Younger patients;

• Patients implanted as soon after the onset of deafness as possible;

• Pre-lingually deafened patients who are in oral programs; and

• Patients who are good lip readers,

generally speaking, are more likely to achieve auditory success than their peers. Other factors may emerge as implant research progresses, but it should be remembered that we have a very strong analog in our collective experience with hearing aid patients who have significant unaided losses, and this experience should guide us. [63]

Of the four factors however, it is interesting to note that each, excepting the last, are almost entirely within our control. [64] With safer electrode designs, the primary remaining danger of implants — the destruction of residual hearing — recedes into the background. As such, we can be more confident in taking the steps these findings imply: in recommending that patients are implanted when younger, as soon after the onset of deafness as possible, and in insuring that implant children are placed in oral education programs.

Some will take exception to my statements, particularly with regard to oral programs: more controversy. Therefore, I wanted to say some things about that.

Education and oral programs

Before the advent of cochlear implants severely and profoundly deaf children were almost exclusively educated in total communication programs. Total communication means that the child is exposed to speech sounds and manual signs simultaneously, while being taught and encouraged to lip read.

Those who are severely or profoundly deaf, even with the best amplification, generally do not gain enough cues from the very limited sound information they receive to enable them to make much use of the speech sounds. Because lip-reading is, to a large degree, a subsidiary skill, best used in supplement to auditory cues, what generally happens in total communication programs is that the primary conduit of information is sign language. I have heard people draw the comparison that teaching the deaf to speak is akin to teaching the blind to paint. Indeed, it may be an exercise inherently fraught with all the difficulty, frustration, and, ultimately, anger that such an exercise could be expected to provoke.

Previous to implants, it was possible to indicate general levels of hearing loss which provided the broad gray line between those children whom we might expect to have some success in an oral program and those for whom the process would be a source of despair. Conventional wisdom was that if the patient could not achieve an aided audiogram which demonstrated anything better than a 50-60 dB loss, then total communication was indicated.

Of course, hearing aids are limited in the degree of amplification they can provide, since if the sound being injected into the ear canal is too loud, it feeds back into the microphone.

Some history is pertinent as well. From the late 1800s to the present a controversy has raged between those who believe oral education is best and those who espouse manual communication. Oral programs are proud of the fact that some children who have unaided losses of 90 to 120 dB have developed spoken language skills. Proponents of manual and total communication insist that this is rare, and therefore for every success story, there are many failures and frustrations. [65] For this reason, they say, all children with severe to profound losses should be immediately started in total communication programs.

Unfortunately, over the course of time, this debate has become very emotionally charged. One could say that the decibel level of the conversation has increased to the point where it has become impossible to take a moderate stance regarding these profound matters which will not be severely criticized.

A strong deaf culture has also emerged which, forgive me, seems in some quarters to have become somewhat reactionary.

For each of these schools of thought, what should be clear is that many of the advantages and disadvantages are entirely personal or familial, and depend on factors which cannot be generalized. In other words, no one can, and therefore no one should claim to be able to choose which solution is best for all deaf children, and this applies whether or not the one espousing the solution is deaf or hearing.

What we must do is to present clear information about each available option, and to let the patients and their families choose, in a consultative, informative, and relaxed atmosphere. Strident arguments will merely increase the pain and confusion of our patients which we should all recognize as among the worst, and yet most easily solved problems consequent to deafness.

A new day

In any case, with the advent of the implant, all previous positions must be re-thought, since implants can often provide a sensation of hearing to those who are otherwise totally deaf. Further, because they do not produce sound in the ear canal, they do not have feedback problems, and so the amount of amplification provided is not limited by this problem. Finally, it may be that implants can offer important subsidiary help to patients who are also using hearing aids, and thus we can use them with a broad range of patients.

In general, given that we can implant patients without destroying residual hearing, what this means is that as implants become more widely accepted, they will necessarily be seen as an option will be tried at some point in the search for a therapeutic regime which offers a given patient the best results.

Once implanted, we must do all that we can to offer the patient the best AI score they can attain, and once that has been done, we must then work with the patient and her or his family in all of the relevant and possible human and auditory dimensions to achieve complete auditory success.