Linear and nonlinear changes in the auditory brainstem response of aging humans.

OBJECTIVE: This experiment was designed to characterize the changes in linear and nonlinear temporal interactions in the aging auditory brainstem of humans using maximum length sequence (MLS) stimulation. METHODS: The MLS technique uses a quasi-random sequence of clicks and silences to determine the linear (linear averaging of single responses) and nonlinear (interactions between pairs or triplets of responses) temporal interactions in the auditory brainstem response (ABR). A group of 30 normal hearing females aged between 11 and 61years were tested. They were divided in three groups: young (<24years); middle (24-41years) and old (>41years). RESULTS: The linear component showed a decrease in wave 5 amplitude with age which could also be related to increased thresholds for higher frequencies. The nonlinear component of the MLS-ABR showed significant differences between middle and old groups for wave 1 and 5 latencies and inter-wave interval (1-5). CONCLUSIONS: The results suggest that the linear and nonlinear components of the MLS-ABR could be more sensitive to changes occurring in the auditory system before any functionally significant changes in hearing status. SIGNIFICANCE: The nonlinear components of the auditory brainstem could be used as an investigative tool to assess early changes in the aging auditory brainstem in young middle-aged women.

Nonlinear properties of otoacoustic emissions in normal and impaired hearing.

Click-evoked otoacoustic emissions (CEOAEs) exhibit nonlinearities in amplitude and time domains. The first objective of this study was to investigate whether there is any correlation between the temporal and amplitude nonlinearities of CEOAEs in normals. Additionally there is evidence that pathology affects the normal cochlear nonlinearity. The second objective was to investigate whether pathology affects the temporal nonlinear components. Conventional and maximum length sequence (MLS) CEOAEs were recorded in normal subjects and in patients with mild hearing loss. The slope of the input-output (I/O) function of the conventional CEOAE measured the amplitude nonlinearity. Two measures of temporal nonlinearity were the magnitude of the suppression that occurs with increase in stimulus rate and the amplitudes of the second and third order temporal interaction components (Volterra slices). The amplitude nonlinearity is well correlated with both the magnitude of the rate suppression and the amplitudes of the Volterra slices. The 'linear' CEOAE amplitude showed no differences between the normal and patient groups but the differences in the Volterra slices were substantial. This suggests that the first sign of damage to the cochlea is that the system becomes more linear. Hence the Volterra slices may provide a sensitive measure of cochlear damage.

Comparison of transient evoked otoacoustic emission thresholds recorded conventionally and using maximum length sequences.

Presenting clicks according to maximum length sequences (MLSs) enables transient evoked otoacoustic emissions (TEOAEs) to be recorded at very high stimulation rates. Despite a decrease in TEOAE amplitude, the very large number of responses obtainable at high rates means that both signal to noise ratio (SNR) and detection sensitivity increase as the click rate increases. This study characterises conventional and MLS TEOAEs near threshold for a group of normally hearing adults. Stimulus presentation rates of 40 clicks/s (conventional) and 5000 clicks/s (MLS) were used. Compared to conventional recordings, the MLS technique enabled smaller responses to be detected, when averaged for the same time and to the same SNR. TEOAE amplitude recorded at detection threshold for MLS responses was 13 dB lower than that recorded conventionally. For each individual, MLS recording also produced clear, repeatable responses at stimulus levels below the detection threshold for conventional TEOAEs. The click level at TEOAE threshold was 12 dB lower for MLS compared to conventional emissions. These results suggest that TEOAE thresholds are not absolute but strongly related to the detection sensitivity of the recording system and physiological noise. The initial growth rates and the shape of input/output functions were found to be similar for the two recording techniques.

Middle ear pathology can affect the ear-canal sound pressure generated by audiologic earphones.

OBJECTIVE: To determine how the ear-canal sound pressures generated by earphones differ between normal and pathologic middle ears. DESIGN: Measurements of ear-canal sound pressures generated by the Etymtic Research ER-3A insert earphone in normal ears (N = 12) were compared with the pressures generated in abnormal ears with mastoidectomy bowls (N = 15), tympanostomy tubes (N = 5), and tympanic-membrane perforations (N = 5). Similar measurements were made with the Telephonics TDH-49 supra-aural earphone in normal ears (N = 10) and abnormal ears with mastoidectomy bowls (N = 10), tympanostomy tubes (N = 4), and tympanic-membrane perforations (N = 5). RESULTS: With the insert earphone, the sound pressures generated in the mastoid-bowl ears were all smaller than the pressures generated in normal ears; from 250 to 1000 Hz the difference in pressure level was nearly frequency independent and ranged from -3 to -15 dB; from 1000 to 4000 Hz the reduction in level increased with frequency and ranged from -5 dB to -35 dB. In the ears with tympanostomy tubes and perforations the sound pressures were always smaller than in normal ears at frequencies below 1000 Hz; the largest differences occurred below 500 Hz and ranged from -5 to -25 dB. With the supra-aural earphone, the sound pressures in ears with the three pathologic conditions were more variable than those with the insert earphone. Generally, sound pressures in the ears with mastoid bowls were lower than those in normal ears for frequencies below about 500 Hz; above about 500 Hz the pressures showed sharp minima and maxima that were not seen in the normal ears. The ears with tympanostomy tubes and tympanic-membrane perforations also showed reduced ear-canal pressures at the lower frequencies, but at higher frequencies these ear-canal pressures were generally similar to the pressures measured in the normal ears. CONCLUSIONS: When the middle ear is not normal, ear-canal sound pressures can differ by up to 35 dB from the normal-ear value. Because the pressure level generally is decreased in the pathologic conditions that were studied, the measured hearing loss would exaggerate substantially the actual loss in ear sensitivity. The variations depend on the earphone, the middle ear pathology, and frequency. Uncontrolled variations in ear-canal pressure, whether caused by a poor earphone-to-ear connection or by abnormal middle ear impedance, could be corrected with audiometers that measure sound pressures during hearing tests.

A study of the acoustic reflex using fast-rate otoacoustic emissions.

Fast-rate otoacoustic emissions (OAEs) were used to determine the inward and outward transmission change produced by the stapedial muscle reflex. The subjects were otologically normal adult volunteers. Satisfactory recordings were obtained from a total of 16 ears. Runs of 16 click-evoked OAEs were recorded at a rate of 4282 clicks/s using the maximum length sequence (MLS) technique, with a contralateral tone presented at 10 dB above the contralateral acoustic reflex threshold during the recording of OAEs 5-12 in each run. The OAEs recorded with the contralateral tone were compared with those recorded without the tone to determine the effect of the acoustic reflex. An analysis of the OAEs in 0.5 kHz frequency bands suggested that the reflex produced significant attenuation of low frequency sound with smaller effects at high frequencies, which included amplification for some subjects. For eight of the 16 ears there was an increase in OAE amplitude with the acoustic reflex for at least one frequency band. At low frequencies (1, 1.5 kHz) the inward transmission change produced by the acoustic reflex appeared to be similar in magnitude to the outward transmission change.

Neonatal otoacoustic emissions recorded using maximum length sequence stimuli.

OBJECTIVE: Maximum length sequence (MLS) stimulation allows transient evoked otoacoustic emissions (TEOAEs) to be recorded at very high stimulation rates. Previous work has focused on recording from normally hearing adult subjects; the aim of this study was to obtain information about emissions recorded using this technique from newborns and to compare these results with those obtained from adults. The feasibility of recording from newborns on the postnatal wards also was addressed. DESIGN: The study comprised two parts. In the first, TEOAEs were collected at 13 stimulation rates from a selected group of babies. The second part of the study comprised only two stimulation rates, a conventional rate of 40 clicks/sec and the maximum MLS rate of 5000 clicks/sec. RESULTS: The neonatal MLS TEOAEs behave in a similar manner to those obtained from adult subjects. The morphology of the waveforms was similar for the conventional and MLS TEOAEs. As the stimulus rate increases, the amplitude of the emission decreases, reaching an approximate plateau by 1000 to 2000 clicks/sec. The absolute reduction in amplitude seen at the high MLS rate is related to the amplitude of the conventional TEOAE but is always approximately the same when expressed as a percentage or proportion of that amplitude. CONCLUSION: The theoretical advantages of speed and sensitivity seen for adult subjects also should hold true for the neonatal population. Although the system used to test was a prototype with none of the refinements found in commercial systems, it was possible to record adequate emissions from a ward-based population of newborns.

Effect of olivocochlear bundle section on evoked otoacoustic emissions recorded using maximum length sequences.

Presenting clicks according to maximum length sequences (MLS) enables transient evoked otoacoustic emissions (TEOAE) to be recorded at very high stimulation rates. As the click rate is increased from 40 clicks/s up to a maximum rate of 5000 clicks/s there is a reduction in TEOAE amplitude that reaches an approximate asymptote at 1500 clicks/s. One hypothesis put forward to explain this MLS 'rate effect' is that ipsilateral efferent activity is involved. To test this hypothesis TEOAEs were recorded from both ears of five patients who had undergone a unilateral vestibular nerve section--a surgical procedure which also entails sectioning the olivocochlear bundle. TEOAEs were recorded conventionally at 40 clicks/s and using MLS stimulation at 5000 clicks/s. Increasing the rate from 40 to 5000 clicks/s was found to reduce the amplitude of the TEOAEs by equivalent amounts in ears ipsilateral and contralateral to a vestibular nerve section as well as in the ears of normal-hearing adults. Since an ear ipsilateral to a vestibular nerve section should have no efferent innervation the hypothesis that efferent activity is the major mechanism involved in the MLS rate effect is rejected. Instead, the possibility that intracochlear processes are the underlying mechanism will now be investigated.

Otoacoustic emissions recorded at high rates in patients with confirmed acoustic neuromas.

Otoacoustic emission testing was carried out on 39 patients with confirmed acoustic neuromas, and satisfactory emissions were recorded from the neuroma ear of 59% of them. A comparison of the patients with and without emissions showed no significant differences in low-frequency or high-frequency hearing loss, optimum speech discrimination score, or canal paresis in the affected ear between the two groups. Emissions were recorded at stimulus rates up to 5,000 clicks/s by using the maximum length sequence (MLS) technique. The decrease in the emission amplitude with increase in click rate (rate suppression) was significantly less than the amount that would be expected from normal subjects for several of the neuroma patients. However, one patient showed normal suppression despite having a large neuroma and no measurable hearing. This would suggest that efferent suppression may not be the only mechanism involved.

The effect of stimulus rate on the contralateral suppression of transient evoked otoacoustic emissions.

A new technique, which uses Maximum Length Sequences (MLSs) to enable the recording of TEOAEs at very fast rates, has been described previously. Such a technique enables rate studies to be carried out and here the effects of stimulus rate on the contralateral suppression of TEOAEs (Collet effect) have been studied. The results show that the Collet effect has virtually disappeared by a click stimulation rate of 2000 clicks/s. The waveform changes caused by the 'adaptation' due to increase in stimulus rate and the 'suppression' due to contralateral stimulation have been evaluated and compared. The comparison indicates that the same mechanism may be involved in both; the increase in stimulus rate leading to an ipsilateral suppression, greater in value but the same in form as the contralateral suppression.

A model to evaluate and maximize hearing aid performance by integrating the articulation index across listening conditions.

A method for integrating the articulation index (AI) across listening conditions was developed and applied to a preliminary model for evaluating and optimizing prescriptions of hearing aid characteristics. The model takes hearing threshold, masking of noise, self-masking of speech, high level cochlear distortion, and the peak-clipping effects of a hearing aid into account. The integrated AI (IAI) across a range of listening conditions is used as a criterion for evaluating a specific hearing aid response characteristic and calculating an optimal frequency-gain characteristic that maximizes the IAI. For a high-frequency hearing loss, the frequency-gain characteristics and IAI's derived from an optimal IAI (OIAI), POGO, and NAL prescriptions are compared for two of listening situations: a quiet setting and a setting with a signal-to-noise ratio of -3 dB. The results predict that, in quiet, the OIAI prescription is not significantly different from the well-established prescriptive procedures such as the POGO and the NAL. For the noise condition, however, the optimal IAI model is predicted to be better in terms of speech intelligibility.

Frequency analysis of the contralateral suppression of evoked otoacoustic emissions by narrow-band noise.

The influence of noise bandwidth on the contralateral masking of click-evoked otoacoustic emissions was investigated by recording emissions from 20 normal subjects with contralateral narrow-band noise, with wide-band noise and without contralateral noise for comparison. Narrow-band noise of five different centre frequencies and three different bandwidths per frequency was used at intensities of 40 and 60 dB SL. Analysis of the emissions showed that the 40-dB SL contralateral noise did not produce a significant amount of suppression. The suppression produced by 60-dB SL noise was spread throughout the frequency range of the emissions, with limited frequency specificity. There was some evidence that the amount of suppression increases with the bandwidth of the noise, particularly for noises centred on 1 and 2 kHz; the wide-band noise produced much greater suppression than any of the narrow bands. There was a large amount of inter-subject variation, which could not be explained by differences in the absolute level of the emissions. One subject with a total unilateral hearing loss in the contralateral ear was also tested, and gave no significant suppression.

High rate otoacoustic emissions.

One of the practical problems of testing young children using evoked otoacoustic emissions (EOAEs) is that there is a need for the child to be quiet for a period of a minute or so. To achieve this can mean test periods of 10 min or longer. A new application of maximum length sequences (MLSs) to EOAEs is described that enables the test to be performed in a few seconds or less. Data from adults and neonates obtained at stimulation rates up to 840 clicks/s are presented and compared with conventionally derived emissions and the implications and applications of this new method are assessed.

Click-evoked otoacoustic emissions: new techniques and applications.

There are practical, clinical circumstances in which a decrease in test time for evoked otoacoustic emissions would be of economic or practical value. In addition, there are circumstances in which it is desirable to obtain more diagnostic information from the evoked emission. This preliminary report concerns predominantly one of these aspects but considers both of them. It is not possible to reduce the test time by simply increasing the stimulation rate because responses and stimuli will overlap and give rise to contaminated responses. However, these limitations can be overcome if a maximum length sequence of clicks is used. These sequences of clicks and silences enable responses of a duration greater than the time between stimuli to be recorded and deconvolved to produce an uncontaminated response. Conventional recordings at stimulus rates of 33/s and 50/s have been taken together with maximum length sequence recordings at rates up to 840/s and the waveforms correlate very well. Although the emission shows some adaptation at the highest stimulation rate this does not prevent a significant reduction in test time.

Isochronic mapping: a preliminary report of a new technique.

A new technique which we have named 'isochronic mapping' is described. The conventional technique of plotting isopotential maps shows little or no contrast for far-field potentials such as the auditory brainstem response (ABR). However, by modifications to the mapping software, the latency values of a peak can be plotted and lines of equal time or isochronic maps can be produced. Data from a normal subject has been obtained for both the compound and derived ABRs and are described in detail. The data are clear but the time delays between earliest and latest projections of a peak are hard to interpret. For monaural click stimulation, wave JV of the ABR projects first to the contralateral side of the head then moves towards the stimulated ear arriving some 0.3 ms later. Possible interpretations of this finding are discussed but further experimentation is needed to develop our understanding of these data.

The choice of stimuli for ABR measurements.

This tutorial describes some of the stimulus choices available to clinicians performing ABR evaluations. The motivation for choosing particular stimuli is described in relation to the clinical question being asked. Thus, emphasis is directed toward the rationales for choosing different stimuli. Data are reviewed only briefly to illustrate these rationales. Finally, some effort has been made to relate clinical ABR results for certain stimulus conditions to underlying physiological processes in the hope that these discussions will lead to a more complete understanding of the stimulus-response relation.

High-frequency monitoring using an electrostimulation audiometer.

Electrostimulation audiometry can provide high-frequency bone-conduction threshold measurements. To evaluate the use of such an audiometer in monitoring hearing, the following areas were investigated. (1) The variability and repeatability of threshold measurements. (2) To see if the instrument could be used on the ward where the patient would have to wear ear plugs to attenuate background noise, the effect of occluding the ears was assessed. (3) The applicability of the instrument to the range of patients in whom monitoring would be appropriate, specifically the relationship between the patient's age and the effective upper frequency limit of the audiometer. The data showed a variability in threshold measurement that is quite comparable with that obtained with conventional audiometry but which became smaller at frequencies of 12 kHz and higher. There was no overall effect on the threshold of occluding the ears, which meant that the instrument can be used to monitor a non-isolated environment such as a ward. The effective cut-off frequency was about 9 kHz for patients aged 55 years, and so the use of an electrostimulation audiometer to monitor high-frequency auditory thresholds is restricted to younger patients.

Electrophysiological responses and stimulus parameters associated with spinal cord stimulation.

Cervical somatosensory-evoked potentials and brainstem-click-evoked potentials were recorded from 10 patients with multiple sclerosis before and after a trial of spinal cord stimulation. An improvement in amplitude of the cervical somatosensory-evoked potential was seen in 7 out of 8 patients, while 3 out of 10 showed a decrease in latency of the brainstem-evoked potential. Stimulus currents were measured during spinal cord stimulation, and those patients requiring high currents to evoke a sensation did not derive clinical benefit from spinal cord stimulation.

Effects of stimulus frequency and intensity on the middle componenets of the averaged auditory eletroencephalic response.

Middle components (latency 8-50 msec) of the averaged auditory electroencephalic response (AAER), evoked by brief duration tone bursts, were recorded from 11 normal-hearing subjects. Latency and amplitude measurements were made on five peaks (Na, Pa, Nb, Pb, and Nc) of the AAER waveforms recorded for 27 experimental conditions: three conditions of stimulus frequency (250, 1000, and 4000 Hz) at each of nine conditions of signal intensity (a no-stimulus control and 10, 20, 30, 40, 50, 60, 70, and 80 dB fe: group thresholds). Latency for each peak decreased with increased stimulus frequency, and it tended to decrease slightly with increases in stimulus intensity. Amplitude input-output characteristics varied with stimulus frequency and response peak. In general, the most linear input-output characteristics occurred for the early peaks and high stimulus frequencies. Characteristics for later peaks and lower frequencies tended to asymptote at moderate stimulus intensities. Between-subject variability was not much greater than within-subject variability for the single event auditory evoked potential (AEP). The variance of the AEP, however, was nearly as great (as much as two-thirds) as the variance of the background EEG, despite the large difference between AEP and background EEG amplitude.