Noise Reduction In Hearing Aids Essay Research

Paper

Recently in a local hearing clinic, a client?s concerns were discussed. ?I?m afraid I won?t like them. My brother in law bought two hearing aids, and he keeps them in a drawer in the kitchen.? While the number of people dissatisfied with their hearing aids hovers around 50%, the hearing aid industry is hard pressed to decrease the number of returns, and increase the average daily use of each aid. In order to accomplish this, hearing aid manufacturers must answer the most often heard complaint: ?It doesn?t work well in noise.?

Unfortunately, a hearing aid will never be able to accomplish the sifting and sorting that is carried out in the human brain. While a person with normal hearing sits in a restaurant, he can distinguish a conversational speech signal that is as little as three decibels greater than the ambient noise. On the other hand, a person with a 30-decibel sensorineural loss might need the speech signal to be 15 or more decibels greater than the ambient noise. The hearing aid?s task is to acoustically or electronically compensate for both the neurological shortcomings of the hearing impaired person and the wide band increase inherent in any basic amplifier.

Acoustic compensation can be carried out in a hearing aid microphone. Most hearing aids today utilize omnidirectional microphones, which pick up sound equally from all directions. This may be beneficial and practical in some cases, as in the completely in the canal (CIC) aid. The CIC aid uses the natural funneling of the auricle in order to focus sound directly into the instrument. Behind-the-ear (BTE) and full concha in-the-ear (ITE) aids lose this anatomical feature, and may benefit from a directional microphone. ?The purpose of using a directional microphone is to focus its sensitivity toward the front of the listener, thereby attenuating or reducing unwanted ?noise? or competition emanating from behind the listener.? (Stach 1998)

Microphone directionality can be accomplished by using a single microphone with two sound inlets. In this mechanical method, the time lag created by a sound entering each inlet is precisely calibrated to cancel out sounds from the sides and back of the microphone as they strike the diaphragm. The Phonak MicroZoom uses an electronic approach. Each MicroZoom contains two omnidirectional microphones. According to the flyer, ?One picks up sound in front of you while a second picks up sound from the sides and rear. A tiny computer chip inside the aid analyzes both sounds. It then enhances the sound from the front, and reduces background sounds from the sides and rear.? (Phonak 1997) In reality, the ?tiny computer? is analyzing the time it takes for a sound to reach each microphone and mathematically decreasing ALL sounds from the sides and rear, speech included.

Directional microphone technology is fallible in one regard: it assumes that all ?noise? is spatially related. Unfortunately, much of the noise present in our environment is omnidirectional. Even though significant benefit can be achieved by ?sound focusing?, a directional aid also must perform in an environment with multiple speakers situated around a listener, combined with noise from high priority directions.

Signal processing, or digital signal processing (DSP), is partly utilized in amplification in order to narrow the focus from ?all? sounds to ?speech? sounds. A linear aid will amplify all frequencies equally, limited to the characteristics of the input / output transducers and the coupler type. On the other hand DSP aids can alter the frequency response of an aid based on their design and/or programming. There are many techniques that can be used which are based on type of loss, but since speech encompasses such a wide frequency range, audiologists need to choose wisely which type of processing to use. Minimally, signal processing for noise reduction is analogous to a continually adjusting graphic equalizer for a home stereo system. Level dependent frequency response (LDFR) hearing aids can also adjust their ?equalizer? settings based on input volume. Along with standard fitting techniques, subjective client opinions are one of the most important factors in decreasing hearing aid returns. In the past several years, technological advances have given audiologists several options.

One such LDFR hearing aid incorporates BILL (bass increase at low levels) circuitry. If it were assumed that most detrimental noise occurs in the low frequencies, then decreasing the low frequency response of a hearing aid at high levels would reduce noise and improve a person?s speech discrimination performance score. This concept seems natural based on the upward spread of masking principle. Since a low frequency sound must travel farther up the basilar membrane, it will mask out a competing higher frequency sound that does not have to travel as far. By reducing low frequency sounds at high levels, a hearing aid could minimize the upward spread of masking, and allow for increased speech reception.

According to one study, while it is reasonable to assume that BILL circuitry would be beneficial in the presence of low frequency noise, ?there is no evidence that listeners with hearing loss exhibit increasing amounts of upward spread of masking with increasing level.? (Bacon, et. al. 1997) Another study shows that while there is a significant difference between unaided performance and BILL aided performance in noise, there was no significant difference between BILL processing and linear processing. (Bess et. al. 1997) Since both studies showed no detriment or reduction in speech scores while using BILL processing, it is reasonable to assume that this type of processing would be beneficial if the client perceives better sound quality as a result.

In contrast, TILL processing or Treble Increases at Low Levels, provides a low frequency signal reduction at low intensities. ?This type of device is based on the premise that hearing-impaired subjects have normal growth of loudness for high level stimuli; therefore, intense sounds should be ?acoustically invisible,? and thus not amplified.? (Bess et. al. 1991) This type of circuit is employed by the K-Amp (Etymotic Research), which is used to decrease loudness discomfort associated with high level, high frequency sounds. ?The K-Amp rationale assumes that most hearing-impaired listeners have greater hearing loss in the high frequencies than in the low frequencies and, therefore, require high-frequency emphasis amplification.? (Valente 1996) Proponents of the TILL circuitry believe that by reducing low frequency sounds at low levels, the hearing aid inherently has a high frequency emphasis, and will therefore promote better speech reception in noise while maintaining a more natural fidelity.

Initially, BILL and TILL seem to be direct opposites, however an insertion gain frequency curve (Bess et. al. 1991) shows that both algorithms provide significant high frequency gain at low levels. With a 50-decibel broad band noise, the BILL hearing aid shows a 10 to 20 dB boost through the low to mid frequencies. At 90 decibels of input, the BILL aid reduces the low frequencies completely, while still providing gain in the high frequencies. The insertion gain frequency curve shows that this type of circuit may not be beneficial for a client with a limited dynamic range. Conversely, the TILL hearing aid shows limited bass boost and significant treble boost with a 50 decibel broad band noise. At 90 decibels of input however, the TILL hearing aid shows a relatively flat response, in retort to the belief that little or no amplification should be necessary at that high intensity.

The BILL and TILL hearing aids are both effective with varying types of hearing losses, but this type of noise reduction is contingent on the fact that noise perception can be changed by varying the frequency response of an amplifier. For the majority of hearing aids, this is only a limited means of changing how well a client can understand speech in noise. Each client however, will have varying degrees of subjective opinions regarding the function of their aid. While acoustic modifications such as venting and mechanical filtering can be made in order to change sound quality, the BILL and TILL hearing aids, once chosen are adjustable only to the limits of their programming.

The PILL circuit (Programmable Increases at Low Levels) affords the clinician more possibilities when adjusting for both noise, and sound quality. ?The most versatile type of ASP (automatic signal processor) comes from PILL circuitry that can provide either a BILL or a TILL response. In order to accomplish this, the instrument must have a minimum of two channel compression.? (Tobin 1997) In effect, the audiologist can choose which type of processing better suits his client, while taking into consideration the subjective needs of his client. Presumably, PILL type hearing aids cost more than BILL or TILL type aids. Adding any programmable feature to a hearing aid generates more expense, and any consumer would need to reap considerable benefit with each increase in price. ?Many users have tried, and failed, with recent approaches at signal processing for improving speech intelligibility in noise. These failures have led many hearing aid users, and their friends, to take a very cautious attitude toward new technology that claims to resolve some of these complaints. Some are convinced that the quality of the amplified signal provided by programmable hearing aids is not significantly better than their present nonprogrammable aids.? (Sandlin 1994) Regardless of these consumer attitudes, a PILL type aid will give a clinician more options when responding to client complaints of poor sound quality, or difficulty with speech intelligibility in noise.

One of the more intriguing recent developments in hearing aid technology promises to deliver a more naturalistic sound quality and unequaled performance in noise. Jonathan Spindel at the University of Virginia has shown advancing growth with a ?magnetic? hearing aid. ?The new device uses a tiny implanted electromagnet attached to the round window of the inner ear to enhance hearing, reduce background noise, and eliminate feedback.? (Source? 1998) Spindel believes that by leaving the human acoustic system intact and unobstructed, the device will be able to create constructive and destructive sound patterns in the cochlea.

The magnetic aid consists of a microphone, a processing unit and an electromagnetic coil, all of which are implanted. For amplification purposes, the aid will generate sound waves that match the ?desired? sound or voice. Combining both signals in phase will have the constructive effect of directly increasing the amplitude of the wave travelling over the basilar membrane. To account for noise reduction, the system is analogous to a FlowMaster muffler (see figure). As shown, the signal processor

generates a sound wave that is out of phase with the incoming signal. The two signals then combine in the cochlea and cancel each other out.

The elimination of feedback is another substantial gain of this type of aid. Since the device does not transmit acoustic energy, magnetic energy at the cochlea generates no feedback potential. ?Our tests to date have shown that the signals produced with our magnetic hearing device are very nearly those of natural acoustic sound.? (Spindel 1998) By not obstructing the normal hearing process and implanting the microphone in the ear, the magnetic hearing aid utilizes a unique feature. The natural resonance of the pinna and external auditory meatus is not impeded so in essence, the device uses our current anatomy as a fully functioning directional microphone.

Although the system is currently being tested on animals with favorable results, human testing will be necessary. The human subject, with a control module, can adjust the aid to reduce undesired sounds and produce maximum benefit. With the advent of miniature computer chips, the magnetic aid allows for an adjustable frequency / intensity analyzer to be implanted in the skull. The noise reduction possibilities for this type of aid are substantial, even if combined with other digital processing techniques.

In conclusion, the advantage of noise reduction techniques in hearing aids is directly related to a reduction in hearing aid complaints, and an increase in client satisfaction. The directional microphone has been a valuable source for noise reduction since early Rock and Roll musicians wanted to reduce feedback during live performances as they got louder and louder. A hearing aid fitted with a directional microphone receives benefit from the assumption that a person will most likely want to attend to a person he is directly facing. In reality, a good portion of our daily communication occurs in less than ideal circumstances. A conversation partner may be just leaving the room when he remembers something important to say, and for normal hearing individuals it is quite commonplace to yell an instruction from one room to another. In some cases, this will defeat the purpose of directivity. Thankfully, manufacturers have responded to these complaints by offering the ability to switch the ?sound focusing? on or off, sometimes with a remote control. (Phonak 1997)

Automatic signal processing in hearing aids is also a valuable advantage for noise reduction. By determining which frequencies are responsible for noise at different intensities, BILL, TILL, or PILL circuits can compress or limit those frequencies from being amplified. Instead of a linear aid, in which all frequencies are amplified equally, these hearing aids reduce noise by limiting frequency bands. Once again, the question becomes a choice as to which frequency bands account for noise. If the hearing aid user is in a room with a noisy air conditioner, low frequency compression may eliminate that noise. If on the other hand the user works in an apiary, high frequency compression during conversation may be a better choice. Frequency range modification is also contingent upon the person?s individual type of loss. For example, common sense dictates that a person with a high frequency hearing loss would probably not benefit from any circuit that provides treble reduction, unless it serves to prevent recruitment.

The magnetic hearing aid offers promises of exquisite noise reduction capability. In theory, the implantable processor will analyze sounds and isolate speech based on a user?s configurations. While the sources don?t reveal how it is possible, or how it might work, the concept is valid. Regardless, each of these noise reduction techniques should be applied on an individual basis, accounting for a person?s nature and degree of loss. Considering the price of inserting additional features into a hearing aid, an audiologist must also balance perceived need with a determination of what actual use might be, and the economical well being of his clients. Perhaps in the future, correct hearing aid selection combined with ever advancing technology will yield the comment, ?Boy, this thing works great in a noisy restaurant!? Alas, that is the nature of the beast.

Bibliography

REFERENCES

Bacon SP, Cook JA, Sammeth CA. (1997). Effect of Low-Frequency Gain Reduction and its Relation to Upward Spread of Masking. Journal of Speech, Language, and Hearing Research. 40(2): 410-422.

Beck LB, Bess FH, Studebaker GA. (1991). Programmable and Automatic Noise Reduction in Existing Hearing Aids. The Vanderbilt Hearing Aid Report II. Parkton, Maryland: York Press; 66-67.

Bess FH, Christensen LA, Hedley-Willams A, Humes LE. (1997). A Comparison of the Benefit Provided by Well-Fit Linear Hearing Aids and Instruments with Automatic Reduction of Low-Frequency Gain. Journal of Speech, Language, and Hearing Research. 40(3): 666-685

Tobin H. (1997). Circuitry Options ? Average Conversation. Practical Hearing Aid Selection and Fitting. Baltimore, MD: Department of Veterans Affairs Rehabilitation Research and Development Service; 24.

Phonak. (1997). The New Standard in Noise-Reduction Technology. MicroZoom. Phonak advertisement.

Sandlin RE. (1994). Dispenser Resistance to Programmable Hearing Aids. Understanding Digitally Programmable Hearing Aids. Needham Heights, MA: Allyn and Bacon; 253.

Stach B. (1998). The Audiologist?s Rehabilitative Tools: Hearing Instruments. Clinical Audiology: An Introduction. San Diego, CA: Singular Publishing Group, Inc.; 469.

Unlisted. (1998). Electromagnetic Hearing Aid. Medical Materials Update. Online 5(11).

(For more information contact: Implantable Device Laboratory, P.O. Box 430, Charlottesville, VA 22908; Tel: 804/924-2050, attn: Jonathan Spindel.)

Valente M. (1996). Amplifiers and Circuit Algorithms of Contemporary Hearing Aids. Hearing Aids: Standards, Options, and Limitations. New York: Thieme Medical Publishers, Inc; p. 189.