Cochlear Hearing Loss and the Current Use of Hearing Aids

Choi, Chul-Hee; Rhee, Chung-Ku

Review

Hearing aids

Korean Journal of Audiology 1999;3(2):111-122.

Cochlear Hearing Loss and the Current Use of Hearing Aids

¹Department of Audiology, Unifversity of Arizona, Tucson, Arizona, USA
²Department of Otolaryngology-Head Neck Surgery, Dankook University College of Medicine, Cheonam, Korea

감각성 난청과 보청기 사용

최철희^¹^, 이정구^²

^¹아리조나대학교 청각학교실
^²단국대학교 의과대학 이비인후-두경부외과학교실

Introduction It is reported that nearly 17 children in 1,000 under 18 years of age and more than two-thirds of individuals above 45 years of age have various degree of hearing loss. The number of Americans with hearing loss comes to approximately 28 million. The largest portion of hearing loss is related to cochlear hearing loss, which involves abnormality, or damage to the cochlear. According to Moore,1) individuals with cochlear hearing loss may exhibit distinctive hearing patterns such as reduced temporal and frequency resolution, reduced hearing sensitivity, abnormal loudness growth, poor sound localization, and difficulty of understanding speech in noise. To overcome these problems, one may consider hearing aids as a main from of treatment. Several hearing aids have been designed to compensate for the perceptual consequences resulting from the cochlear damage. Selecting a proper hearing aid is important for people with cochlear hearing loss. However, it is not to say that the use of hearing aids is a panacea for the cochlear hearing loss. Even with the use of hearing aids, it is difficult for the damaged cochlea to be restored to its previous status. We discuss major perceptual consequences resulting from cochlear damage, as well as how the current hearing aids are designed. Finally, this paper reviews studies on the effectiveness of current hearing aids. Cochlear Hearing Loss Perceptual Consequences Resulting from Cochlear Damage The cochlea is the most important organ of an ear. The basilar membrane (BM) is of primary concern because it is more sensitive to vibrations of sound than are other structures.2) The BM is narrow (less than 0.1 mm) and thick at the base while it is wide (about 0.5 mm) and thin at the apex. According to Durrant,3) there is “a systematic variation in the width of the basilar membrane from the base to the apex of the cochlea.” The motion of the BM in response to sound takes the form of a travelling wave. Bekesy originally described the travelling wave. Based on Bekesy's theory, the amplitude of the wave increases first and then decreases rather abruptly.4) Fig. 1-1 shows the instantaneous displacement of the BM as a function of distance from the stapes. This graph shows two successive waves in time, moving from left to right, in response to a 200-Hz sinusoid. Furthermore, Moore4) reported that sounds of different frequencies produce maximum displacement at different locations on the BM. For example, high frequency sounds produce maximum amplitude at the basal end of the BM while low frequency sounds produce it at the apical end. Recent studies of BM vibration have reported several important findings. First, the BM is sharply tuned and the sharpness of tuning of the BM depends on the physiological condition of the animal.4) Fig. 1-2 shows the turning curves of the BM. In Fig. 1-2, the BM is sharply tuned when the action potential (AP) threshold is low, indicating good physiological condition. Secondly, the BM vibration is linear for low input sound levels but nonlinear for high input levels. The magnitude of the BM response does not grow directly in proportion with the magnitude of the imput. Thirdly, the BM respones varies with different types of sounds. The BM response to a single click is more complex than that of sinusoid. On the basilar membrane lies the hearing organ called the organ of Corti. The organ has two types of hair cells- outer and inner hair cells. Both cell types have hairlike structures called stereocilia on the top. However, the outer and inner hair cells differ in the shape of their cell bodies and their numbers as well as in the configurations of their bundles of stereocilia.3) In the human cochlea, there are approximately three to five parallel rows of 12,000 to 15,000 outer hair cells and a single row of about 3,000 inner hair cells.2) According to Durrant,3) while the stereocilia of the inner hair cells form an almost cotinuous line along the organ of Corti, with the sterocilia of each forming a shallow crescent. On the outer hair cells, the stereocilia are arranged in a distinctive “W” shape. Damage to the hair cells in the mammalian ear results in mammalian ear results in cochlear hearing loss.5)6) Unfortunately, once the hair cells are destroyed, they do not naturally regenerate themselves. However, when artificially stimulated to regenerate, the hair cells of some species, such as sharks and birds, are capable of regeneration.6)7) According to Audiology today,7) Oberholtzer reported, that after destruction of the hair cells in chickens, supporting cells such as Deiters', and Claudius's cells began to divide and grew into hair cells when treated with forskolin. In spite of the possibility of regeneration of the hair cell in birds, application of this research to mammalian ears is still in dispute. Damage to the hair cells may result from either endogenous or exogenous causes.2) The endogenous causes consist of prenatal and postnatal factors. Prenatal factors include hereditary disorders such as autosomal dominant/recessive inheritance, X-linked hearing loss；the Rh factor；cerebral palsy；viral infection (rubella)；and anoxia. Postnatal factors include otitis media, meningitis, labyrinthitis, and some viral infections (measles, mumps, chicken pox, and influenza). Common exogenous causes are drug abuse (quinine, kanamycin, neomycin, viomycin, and dihydrostreptomycin), noise exposure, presbycusis, and head trauma. Liberman et al.8) have shown structural changes in the damaged cochlea produced by noise exposure or treatment with ototoxic drugs. In some cases, the inner hair cells were damaged while the outer hair cells remained nearly normal. In other cases, the outer hair cells were damaged while the inner hair cells remained apparently intact. Fig. 1-3 shows those cases exhibiting partial loss of the outer hair cell (OHCs) with intact the inner hair cells (IHCs). This change is typical of ototoxic durges, but is less typical of noise exposure. A normal tuning curve (solid) is compared with the abnormal tuning curve (dotted) associated with this damage. The abnormal tuning curve appears to have two sections, an elevated tip, and a hypersensitive tail. Fig. 1-4 shows the cases which demonstrate total loss of OHCs and intact IHCs. This pattern is typical of ototoxic drugs. The abnormal tuning curve is bowl-shaped. Fig. 1-5 shows those cases in which both the IHCs and OHCs are severely damaged. In this case, the entire tuning curve is shifted upwards and is missing the sharp downward tip, which is present in the normal curve. The reduced sensitivity to the higher frequencies results from damage to the IHCs. Fig. 1-6 shows the cases in which the IHCs are moderately damaged but the OHCs are minimally damaged. The tuning curve is nearly normal, but the whole curve is shifted upwards by about 40 dB. Moore1) summarizes the results as follows：“The OHCs are responsible for the sharp tips of the tuning curve. When the OHCs are damaged, the sharp tip becomes elevated, or may disappear together. Therfore, the damage of the OHCs result in a threshold elevation around the tip of the tuning curve of 40-50 dB. However, the IHCs are sensitive to the tail of the tuning curve, whether or not the OHCs are damaged. Damage to the IHCs result in an overall loss of sensitivity.” After reviewing recent studies about the functions of hair cells, Moore9) summarizes physiological consequences of the IHCs and OHCs damage as follows：Damage to the IHCs produce several consequences such as less efficient transduction of mechanical vibration into neural activity, loss of sensitivity, reduced information flow in the auditory nerve, and no transduction of activity at some regions of the BM. Damage to the OHCs produces adverse consequences such as reduced amplitude of vibration on the BM in response to weak sounds, reduced sharpness of tuning on the BM, and more linear input-output functions on the BM. These physiological consequences are related to perceptual problems such as reduced hearing sensitivity, abnormal loudness growth, and difficulty of understanding speech in noise.1) Several hearing aids have been designed to compensate for the perceptual problems. Linear aids have been used for reduced hearing sensitivity. Compression and noise reduction hearing aids have been used to compensate for both abnormal loudness growth and difficulty in understanding speech in noise. Loss of Hearing Sensitivity and Linear Hearing Aids One of the most obvious signs displayed by people with cochlear loss is difficulty of hearing weak sounds, referred to as audiometric loss.10) Audiometric loss presents as increased hearing thresholds in dB at various frequency on the audiogram. This degree of hearing loss may be classified as mild, moderate, severe, or profound. According to Moore,11) this problem results from reduced frequency selectivity, which refers to the reduced ability of the auditory system to separate or resolve the frequency components in a complex sound. The loss of hearing sensitivity may be illustrated using the articulation index (AI), often referred to as the audibility or speech intelligibitlity index. This index has been used for assessing hearing sensitivity loss, as well as for measuring the potential effectiveness of hearing aids. The AI has been used to quantify how much of the speech spectrum is audible to listeners.11) The AI is based on importance that each frequency region can contribute to speech intelligibility. Pavlovic12) developed the original computation method, which Mueller & Killion13) later simplified. The “couni-the-dot method” uses 100 dots, which are distributed among the different frequencies according to their importance in understanding speech. The number of dots at a particular frequency indicates the relative importance of that frequency in speech comprehension (Fig. 2-1). Mueller & Killion14) suggested that the AI can be used as follows：(1) to predict the amount of the hearing sensitivity loss for normal speech from the unaided audiogram, (2) to predict the benefit that hearing aids can provide, and (3) to compare effects of hearing aids. Fig. 2-2 shows that AI score for unaided (20%), aided (79%), and target aided (74%) conditions. The AI score (20%) for the unaided condition indicates that the patient is unable to hear 80% of the speech spectrum. Linear hearing aids, often called peak clipping hearing aids, have been used to compensate for reduced hearing sensitivity. As illustrated in Fig. 2-3, an increase in the level of the input signal results in an even increase in output level until the saturation level is reached. Gain refers to the difference in dB between the input level and the output level. When the hearing aid exceeds its maximum output limits, peak clipping occurs in the hearing aid. Peak clipping results in numerous forms of distortion. The amount of distortion increases as the input or gain of the hearing aid increases beyond the clipping threshold. Increase in distortion has been reported to reduce speech intelligibility and sound quality.15) Linear hearing aids have also been used for noise reduction.16-18) In this case, linear hearing aids with a high-pass filter have been utilized to decrease the intensity level of environmental noises, which are generally low-frequency energy. However, this type of hearing aid is disadvantageous for individuals with low frequency hearing loss since some environmental noises, such as cafeteria and multitalker noises, have substantial energy in the middle and high frequencies.18) The linear aids with high-pass filters are not useful for reducing the effect of these noises. In summary, linear hearing aids are often ineffective in providing the amplification needed to accommodate the auditory deficit without distortion. Abnormal Loudness Growth and Compression Hearing Aids To solve the problem of distortion, compression hearing aids may be used. However, compression hearing aids were originally designed to compensate for loudness recruitment, not for the loss of hearing sensitivity. Generally, as the intensity of sound increases, so does the sensation of loudness for listeners with normal hearing. However, individuals with cochlear hearing loss experience unusually repid growth of loudness as the intensity of sound increases.2) This phenomenon, known as recruitment, is related to reduction in dynamic range, the difference between hearing thresholds and loudness discomfort level (LDL). Moore (1997) explained that recruitment results mainly from loss of the normal compression that occurs on the BM. In other words, damage to OHCs results in steeper-than-normal input-output functions. Killion & Fikret-Pasa3) conducted an interesting study illustrating three types of hearing loss based on loudness growth. Fig. 3-1 shows the loudness growth resulting from a mild-moderate sensitivity loss of 40 dB, called Type Ⅰ hearing loss. It illustrates complete recruitment. Individuals with Type Ⅰ hearing loss have loss of sensitivity for soft sounds, but no loss of hearing sensitivity for loud sounds, as illustrated in Fig. 3-1. In this example, weak sounds below 40 dB HL are inaudible, but intense sounds above 80 dB HL are perceived normally. Above 40 dB HL, the sensation of loudness is gradually regained with increasing intensity. For individuals with Type Ⅰ hearing loss, other aspects of hearing such as frequency resolution, loudness resolution, ability to understand speech in noise, and ability to perform in a musical group are normal or near normal at the higher frequencies. Individuals with Type Ⅱ hearing loss require more gain for soft sounds, but less gain (8 or 10 dB) for loud sounds. Fig. 3-2 shows the partial recruitment which defines a Type Ⅱ hearing loss. A moderate-severe loss of 60 dB is illustrated. In this Fig., weak sounds below 60 dB HL are inaudible, while intense sounds above 80 dB HL are within 10 dB of normal. Fig. 3-3 illustrates the loudness growth curve of a Type Ⅲ hearing loss. A severe-to-profound hearing loss of 70 dB is depicted. As illustrated by the graph, the dynamic range for intelligible speech in difficult listening environments is very narrow (5-10 dB on the HL dial). Compression hearing aids were originally designed to accommodate the loudness recruitment found in individuals with cochlear hearing loss. Compression hearing aids essentially compress sounds or speech from a large dynamic range into a small dynamic range. In this type of hearing aid, the output decreases automatically as the input level increases. Compression hearing aids can exist in a wide variety of forms.15) Compression hearing aids can be classified as input, output, or gain controlled, according to the manner in which the circuit is designed. They may also be classified by their circuit characteristics, which may include automatic gain controls (AGC), syllabic compression, compression limiting, wide dynamic range compression, frequency dependent compression, curvilinear compression, and multiple channel compression. Fig. 3-4 illustrates the characteristics of a compression hearing aid relative to input-output function.19) The compression threshold refers to the input level at which the compressor starts to operate. The compression range indicates the range of input levels over which the compressor operates. The compression ratio refers to the ratio of a change in input level to the corresponding change in output. Mueller and Killion14) explained how compression operates in individuals with three different types of cochlear hearing loss. Fig. 3-5 shows a Type Ⅰ hearing loss with and without the use of linear hearing aid with peak cliping. The linear hearing aid operates well enough for quiet sounds. However, the output of the linear hearing aid is abnormally loud between 40 dB HL and 90 dB HL input levels. For example, at input levels between 60 dB and 70 dB HL the perceived loudness is 25 dB above normal levels. Fig. 3-6 shows the aided loudness function using a wide-dynamic range compression that has a 2.2：1 compression ration, with a lower compression threshold at 30 dB HL and higher threshold at 75 dB HL. Fig. 3-7 illustrates the aided loudness functions of two different types of compression circuits, curvilinear compression and 3：1 compression. Fig. 3-8 shows both the unaided and aided loudness functions for several types of compression hearing aids in the case of the Type Ⅱ hearing loss. For this type of hearing loss, a hearing aid should provide 35 dB to 40 dB of gain for quiet sounds and 10 dB of gain for intense sounds. In Fig. 3-8, the 3：1 compression ratio permits better loudness matching than either the 2.2：1 compression ratio or the curvilinear or 3：1 circuits. As illustrated in the above examples, the ability of compression hearing aids to compensate for loudness recruitment provides significant benefits for those with Type Ⅰ and Type Ⅱ hearing losses. Unfortunately, these benefits are somewhat offset by some undesirable side effects such as distortion of the temporal envelope of sounds, and pumping or breathing sounds.9) Furthermore, although compression hearing aids have improved the overall audibility of speech, speech recognition in noise is still problematic. Since compression hearing aids are not able to separate the speech signal from noise but only reduce the relative level of noise, the do not function well for improving speech recognition in noise. 20) Difficulyt of Understanding Speech and Noise Reduction Hearing Aids One of the complaints of individuals with cochlear hearing loss is loss of clarity, which refers to difficulty in understanding speech in noisy environments.10)21)22) The problem of speech comprehension in noise is frequently described as “I can hear what people say, but I can't understand them”.10) The HINT (Hearing in Noise Test) abd SIN (Speech in Noise) tests are used to measure the proble. Results are expressed as the signal-to-noise ratio (SNR) in dB required for 50% correct recognition of words.10) In order to achieve normal speech intelligibility, people with cochlear hearing loss may require SNR 16 dB or more.4) There are several viowpoints regarding the effect of noise on speech comprehension. Killion10) attributes the problem to hair cell damage. Humes23) provides an audibility-based explanation, claiming that “noise masks speech in the lower frequencies and reduces audibility of high frequency consonants.” Kuk24) claims this problem is due to decreased frequency and decreased temporal resolutions. Sammeth18) and Preves25) assert that it comes from upward spread of masking, suggesting that “the effect of low frequency noise to high frequency speech is greater than that of high frequency noise to low frequency speech.” Hearing aids with a directional microphone or mulit-microphone array have been used for noise reduction. Adaptive hearing aids and directional hearing aids differ in the manner in which they process signals. The directional microphone acts on the signal “before it enters the hearing aid's signal processing ports, whereas the adaptive filter is placed with the signal processing path”.20) Assuming that speech signals mostly originate in front, but that noise comes from all directions, the use of a directional microphone can reduce some noise interference by picking up input signals in the front only. With one directional microphone, however, the speech energy is also reduced.25) To solve the low sensitivity problem of one directional microphone, a multimicrophone array has been used to improve the intensity as well as the directionality of a hearing aid.26) In the multimicrophone array, directional sensitivity is increased by “subjecting the outputs of two or more spatially separated microphone to a combination of filtering, time delay, and amplitude weighting”.15) Fig. 4-1 illustrates directional pattern with a two-microphone array and three-microphone array. One hearing aid manufacturee, Phonak, has recently introdcued a programmable heairng aid, the PiCS AudioZoom, which has two microphones placed in the same hearing aid. With the AudioZoom program, sounds in front of listeners are amplified more than sounds from other directions. The directional microphone array is an example of true noise reduction technology because it theoretically reduces the level of noise relative to the level of the signal. Assuming that the theories regarding spatial distribution of signal and noise are valid, it should produce a measurable increased of signal-to-noise ratio (SNR). It has been demonstrated that directional hearing aids based on the array technology can improve the SNR as much as 7.5 dB.27) Evaluation of the Effectiveness of Hearing Aids Many studies have been undertaken to determine the benefit of hearing aids. Dillon28) reviewed published research about the effect of several types of compression hearing aids. The results indicated that compression circuits did not offer better intelligibility than linear hearing aids for conversational-level speech in quiet. However, if the input level was decreased, some types of compression provided intelligibility superior to that available from linear hearing aids. In addition, compression hearing aids provided significant intelligibility advantages in broadband noise. According to Preves,25) although several studies have shown better improvement in speech intelligibility using compression hearing aids compared to linear hearing aids, other studies have reported improvement only with multichannel, but not with single-channel, compression hearing aids. However, multiband compression circuits have failed to demonstrate significant improvement in speech intelligibility over single channel compression and linear hearing aids.25) Schum29) studied speech recognition in a background of cafeteria noise for elderly listeners using a linear hearing aid and hearing aids featuring one of the following；(1) high frequency fixed filter, (2) a directional microphone, (3) automatic signal processing (ASP) circuit, and (4) a noise reduction circuit. Sixteen elderly people from 60 to 77 years old with symmetrical sensorineural hearing loss were used as subjects. The CID W22 monosyllabic word lists were used as the speech stimuli. The speech material was presented at 0 degree aximuth and the cafeteria noise was presented at 180 degree azimuth via two loudspeakers. The speech material was presented at 67 dB SPL and the sound pressure level for the cafeteria noise was varied in 2 dB steps, beginning with a SNR of ＋12 dB. The mean SNR necessary for a 50% correct recognition point was obtained. According to Schum's report,29) all four hearing aids provided a similar magnitude of benefit over the linear hearing aid. A large inter-subject variability was also reported. Gordon-Salant30) assessed the effectiveness of and adaptive frequency respones hearing aid (AFR) and a linear hearing aid on speech perception in noise. It was reported that the hearing aid with AFR had significant high ratings on speech recognition, but not on any other ratings. It was shown that the Revised Speech Perception in Noise test (R-SPIN) provided the most sensitive measure for revealing differences in speech recognition, whereas the qualitative judgment tasks were not sensitive to differences in different hearing aids due partially to wide inter- and intra-subject variability. Valente26) investigated the effectiveness of the dual microphones of the Audio-Zoom using the Profile of Hearing Aid Benefit (PHAB) and Abbreviated Profile of Hearing Aid Benefit (APHAB). Fifty adult subjects with mild to moderately severe sensorineural hearing loss wore binaural behind-the-ear hearing aids and used a remote control to switch between the two programs (“basic” and “party”) and two microphone conditions (omnidirectional and directional). Speech recognition scores in noise were measured in terms of SNR (the difference in dB between a signal level required for 50% correct speech and the noise level) using the Hearing in Noise Test (HINT). It was reported that there was and average improvement in SNR oF 7.4 to 8.5 dB when using the directional microphone (AudioZoom), whereas there was no significant improvement in SNR between the two programs (“basic” and “party”). Assuming that a 1 dB improvement in SNR leads to and improvement in speech recognition scores of 8.5% on the HINT, the obtained SNR improvement may be enhanced 62% to 72% in sentence intelligibility. They reported that the PHAB and APHAB scores for speech intelligibility in background noise with reduced cues as well as scores for aversiveness to sounds were significantly higher than those in the established nroms. In addition, they found that 76% of the subjects preferred the AudioZoom to their current hearing aids. These results, however, are inconsistent with the outcomes of Schum's study about the benefit of the noise reduction hearing aids. The inconsistent results may be attributable to methodological differences. Conclusion In summary, a selective review of the literature on the evaluation of the effectiveness of hearing aids suggests that it is difficult to determine the benefit of hearing listeners' rating alone because of (1) intra- and inter-judge variability, (2) methodlogical differences, (3) the use of different hearing aids, and (4) the use of different speech and noise stimuli. Although several studies demonstrated that subjects were able to recognize a difference in performance among different hearing aids, other studies contradict these findings. Listener ratings may be affected by extraneous factors such as memory, personality, education, age, health, mood, and experience with hearing aids. In addition, the listener ratings have poor test-retest reliability, putting the validity, objectivity, and efficiency of these studies in question. Despite disagreement among researchers regarding the effectiveness of hearing aids in increasing the signal-to-noise ratio, it is agreed that current hearing aids are able to compensate for loss of sensitivity, or audibility. It is hoped that more advanced technologies, such as digital signal processing and directional microphones, will be able to solve the problems associated with reduced speech comprehension in noise.
REFERENCES
1) Moore BCJ. Perceptual consequencies of cochlear hearing loss and their implications for the design of hearing aids. Ear and Hearing 1996;17(2):133-61. 2) Martin FN. Introduction to Audiology. New Jersey: Prentice-Hall, Inc., 1994. 3) Durrant JD, Lovrinic JH. Bases of hearing sciences. Baltimore: Williams & Wilkins Co., 1995. 4) Moore BCJ. Introduction to the psychology of hearing. San Diego: Academic Press, 1997. 5) Moore BCJ. Perceptual consequencies of cochlear damage. Oxford: Oxford University Press, 1995. 6) Salvi RJ, Chen L, Trautwein P, Powers N, Shero M. Hair cell regeneration and recovery of function in the avian auditory system. Scandinavian Audiology 1998;27(Suppl. 48):7-14. 7) A Publication of Starkey Technical Servieces. The Compression Handbook: An overview of the characteristics and application of compression amplification, 1994. 8) Liberman MC, Dodds LW, Learson DA. Structure function correlation in noise-damaged ears: A light and electron-microscopic study. In: RJ Salvi, D Henderson, RP Hamernik, V Colletti. (Ed). Basic and Applied Aspects of Noise-Induced Hearing Loss. New York: Plenum, 1986;163-76. 9) Moore BCJ. Cochlear hearing impairment and the design of hearing aids. The Newsletter of Acoustical Society of America 1998;8:4-7. 10) Killion MC. SNR loss: “I can hear what people say, but I can't understand them.” The Hearing Review 1997;4(12):8-14. 11) Popelka GR. Computer technology and hearing aids. In: RE Sandlin. (Ed). Handbook of hearing aid amplification: Vol. I, theoretical and technical considerations. San Diego: Singular Publishing Group, Inc., 1995;239-63. 12) Pavlovic C. Articulation index predictions of speech intelligi-bility in hearing aid selection. Asha 1988;30(6/7):63-5. 13) Killion MC, Fikret-Pasa S. The 3 types of sensorineural hearing loss: loudness and itelligibility considerations. The Hearing Journal 1993;46(11):31-6. 14) Mueller HG, Killion MC. An easy method for calculating the articulation index. The Hearing Journal 1992;45(9):14-7. 15) Fortune T. Amplifiers and circuit algorithms of contemporary hearing aids. In: M. Valente. (Ed). Hearing aids: standards, options & limitations. New York: Thieme Medical Publishers, Inc., 1996;157-209. 16) Schum DJ. Speech understanding in background noise. In: M Valente. (Ed). Hearing aids: standards, options, & limitations. New York: Thieme Medical Publishers, Inc., 1996;368-406. 17) Levitt H. Future direction in hearing aid research. Journal of Speech-Language Pathodology and Audiology Monogr 1993;1(Supp. l):107-24. 18) Sammeth CA, Ochs MT. A review of current “noise reduction” hearing aids: rationale, assumptions, and efficacy. Ear and Hearing 1991;12(Suppl. 6):116S-24S. 19) Dillon H. Compression in hearing aids. In: RE Sandlin (Ed). Handbook of hearing aid amplification: Volume I, theoretical and technical considerations. San Diego: Singular Publishing Group, Inc., 1995;121-45. 20) Bachler H. Vonlanthen A. AudioZoom signal precessing for improved communication in noise. Phonak Focus 1995. p.18. 21) Verschuure J, Dreschler WA. Present and future technology in hearing aids. Journal of Speech-Language Pathology and Audiology Monogr 1993;Suppl. 1:65-73. 22) Fabry DA, Walden BE. Noise-reduction hearing aids: what is the fate of ART (adaptive response technology)? ASHA 1990;32(June/July):48-51. 23) Humes LE. Understanding the speech-understanding problems of the hearing impaired. Journal of the American Academy of Audiology 1991;2:59-69. 24) Kuk FK, Tyler RS, Stubbing PW, Bertschy MR. Noise reduction circuitry in ITE instruments. The Hearing Instruments 1989;40(7):20-6. 25) Preves DA. Approaches to noise reduction in analog, digital, and hybrid hearing aids. Seminars in Hearing 1990;11(1):39-67. 26) Valente M, Fabry DA, Potts LG. Recognition of speech in noise with hearing aid using dual-microphone. Journal of the American Academy of Audiology 1995;6(4):440-9. 27) Soede W, Berkhout AJ, Bilsen FA. Development of a new directional instrument based on array technology. The Journal of the Acoustical Society of America 1993;94:785-98. 28) Dillon H. Tutorial Comression? Yes, But for low or high frequencies, for low or high intensities, and with what response time? Ear and Hearing 1996;17(4):287-307. 29) Schum DJ. Noise reduction strategies for elderly, hearing impaired listeners. Journal of the American Academy of Audiology 1990;1(1):31-6. 30) Gordon-Salant S, Sherlock LP. Performance with an adaptive frequency response hearing aid in a sample of elderly hearing impaired listeners. Ear and Hearing 1992;13(4):255-62.