Robust estimation of synchronized spontaneous otoacoustic emission via singular value decomposition and optimal shrinkage

We investigate matrix signal processing techniques for estimating synchronized spontaneous otoacoustic emission (OAE) in noise. Responses to repeated clicks are first stored in a matrix, and singular value decomposition is either applied in the time domain or the frequency domain after constructing a Hankel matrix at every frequency. The singular values are subject to optimal shrinkage (OS) which maximizes the signal-to-noise ratio. Human OAE data were analyzed, and the Hankel matrix method outperforms the time-domain OS method in synchronized spontaneous otoacoustic emission estimation, but not in the estimation of transient-evoked otoacoustic emission. Reasons for the performance discrepancy are discussed.

Model-based prediction of otoacoustic emission level, noise level, and signal-to-noise ratio during time-synchronous averaging

Although averaging is effective in reducing noise, its efficiency rapidly decreases beyond several hundred averages. Depending on environmental and patient noise levels, several hundred averages may be insufficient for informed clinical decision making. The predictable nature of the otoacoustic emission (OAE) and noise during time-synchronous averaging implicates the use of predictive modeling as an alternative to increased averaging when noise is high. Click-evoked OAEs were measured in 98, normal-hearing subjects. Average OAE and noise levels were calculated for subsets of the total number of averages and then fit using variants of a power function. The accuracy of the models was quantified as the difference between the measured value and model output. Models were used to predict the OAE signal-to-noise ratio (SNR) for a criterion noise level. Based on predictions, the OAE was categorized as present or absent. Model-based decisions were compared to decisions from direct measurements. Model accuracy improved as the number of averages (and SNR in the case of OAEs) from which the model was derived increased. Model-based classifications permitted correct categorization of the OAE status from fewer averages than measurement-based classifications. Furthermore, model-based predictions resulted in fewer false positives (i.e., absent OAE despite normal hearing).

A point-wise artifact rejection method for estimating transient-evoked otoacoustic emissions and their group delay

Due to its low intensity, measurement of transient-evoked otoacoustic emission (TEOAE) requires repeated stimulation. When any acoustic artifact occurs, an entire click interval is typically abandoned. Here, a point-wise artifact rejection strategy is proposed, and it partially preserves the data when artifacts occur in an interval. At the noisiest setting (-46 dB signal-to-noise ratio) the proposed strategy retains four times more data and thereby reduces the root mean square signal estimation error by over 60%. Consequently, the group delay can be calculated more accurately. These findings might facilitate TEOAE measurement at home or in other noisy environments in the future.

Measuring Distortion-Product Otoacoustic Emission With a Single Loudspeaker in the Ear: Stimulus Design and Signal Processing Techniques

The distortion-product otoacoustic emission (DPOAE) is a backward propagating wave generated inside the cochlea during the wave amplification process. The DPOAE signal can be detected rapidly under relatively noisy conditions. In recent years, the earphone industry demonstrated interest in adopting DPOAE as an add-on feature to make their product “intelligent” of inner-ear status. However, a technical challenge remains to be tackled—the loudspeaker in an earphone generates its own cubic distortion at the same frequency as DPOAE. Unfortunately, the intensity of loudspeaker distortion is typically comparable to that of the DPOAE, if not higher. In this research, we propose two strategies, namely compensation and cancellation, to enable DPOAE measurement with a single loudspeaker. The compensation strategy exploits the part of the growth function of the loudspeaker distortion which is almost linear, and thus suppresses the distortion it generates while retaining a larger portion of DPOAE in the residual signal. The cancellation strategy utilizes a one-dimensional Volterra filter to remove the cubic distortion from the loudspeaker. Testing on normal-hearing ears shows that the compensation strategy improved the DPOAE-to-interference ratio by approximately 7 dB, resulting in a cross-correlation of 0.62 between the residual DPOAE level and the true DPOAE level. Meanwhile, the cancellation strategy directly recovered both the magnitude and the phase of DPOAE, reducing the magnitude estimation error from 15.5 dB to 3.9 dB in the mean-square sense. These pilot results suggest that the cancellation strategy may be suitable for further testing with more subjects.

Systematic Audiological Assessment of Auditory Functioning in Patients With Parkinson's Disease

Purpose Alterations in primary auditory functioning have been reported in patients with Parkinson's disease (PD). Despite the current findings, the pathophysiological mechanisms underlying these alterations remain unclear, and the effect of dopaminergic medication on auditory functioning in PD has been explored insufficiently. Therefore, this study aimed to systematically investigate primary auditory functioning in patients with PD by using both subjective and objective audiological measurements. Method In this case-control study, 25 patients with PD and 25 age-, gender-, and education-matched healthy controls underwent an audiological test battery consisting of tonal audiometry, short increment sensitivity index, otoacoustic emissions (OAEs), and speech audiometry. Patients with PD were tested in the on- and off-medication states. Results Increased OAE amplitudes were found when patients with PD were tested without dopaminergic medication. In addition, speech audiometry in silence and multitalker babble noise demonstrated higher phoneme scores for patients with PD in the off-medication condition. The results showed no differences in auditory functioning between patients with PD in the on-medication condition and healthy controls. No effect of disease stage or motor score was evident. Conclusions This study provides evidence for a top-down involvement in auditory processing in PD at both central and peripheral levels. Most important, the increase in OAE amplitude in the off-medication condition in PD is hypothesized to be linked to a dysfunction of the olivocochlear efferent system, which is known to have an inhibitory effect on outer hair cell functioning. Future studies may clarify whether OAEs may facilitate an early diagnosis of PD.

A long-term follow-up study on otoacoustic emissions testing in paediatric patients with severe malaria in Gabon

BACKGROUND

In a previous study, severe and cerebral malaria have been connected with acute cochlear malfunction in children, demonstrated by a decrease of transitory evoked otoacoustic emissions (TEOAEs) reproducibility. This study aims to determine whether cochlear malfunction persists for 4 years after recovery from severe malaria in a subset of the previous study's collective. Follow-up TEOAEs were performed on site (CERMEL, Hôpital Albert Schweitzer, Lambaréné, Gabon) or at the participants' homes; 33 out of 90 participants included in the initial investigation by Schmutzhard et al. could be retrieved and were re-examined, 31/33 could be included. Of the 57 missing participants, 51 could not be contacted, 1 had moved away, 4 refused to cooperate, and 1 had died.

METHODS

As in the initial investigation, participants of this prospective follow-up study were subjected to TEOAE examination on both ears separately. A wave correlation rate of > 60% on both ears was considered a "pass"; if one ear failed to pass, the examination was considered a "fail". The results were compared to the primary control group. Additionally, a questionnaire has been applied focusing on subsequent malaria infections between the primary inclusion and follow-up and subjective impairment of hearing and/or understanding.

RESULTS

The cohort's mean age was 9 years, 14 children were female, 18 male. 31 had been originally admitted with severe, one with cerebral malaria. 83.8% of participants (n = 26) presented with a TEOAE correlation rate of > 60% on both ears (the cut-off for good cochlear function); in the control group, 92.2% (n = 83) had passed TEOAE examination on both ears. Recurrent severe malaria was associated with a worse TEOAE correlation rate. Age at infection and gender had no influence on the outcome.

CONCLUSIONS

Cochlear malfunction seems to be persistent after 4 years in more than 16% of children hospitalized for malaria. In a healthy control group, this proportion was 7.8%. Yet, the severity of the initial TEOAE-decrease did not predict a worse outcome.

Adaptive filter based two-probe noise suppression system for transient evoked otoacoustic emission detection

Transient otoacoustic emission (TEOAE) is a method widely used in clinical practice for assessment of hearing quality. The main problem in TEOAE detection is its much lower level than the level of environmental and biological noise. While the environmental noise level can be controlled, the biological noise can be only reduced by appropriate signal processing. This paper presents a new two-probe preprocessing TEOAE system for suppression of the biological noise by adaptive filtering. The system records biological noises in both ears and applies a specific adaptive filtering approach for suppression of biological noise in the ear canal with TEOAE. The adaptive filtering approach includes robust sign error LMS algorithm, stimuli response summation according to the derived non-linear response (DNLR) technique, subtraction of the estimated TEOAE signal and residual noise suppression. The proposed TEOAE detection system is tested by three quality measures: signal-to-noise ratio (S/N), reproducibility of TEOAE, and measurement time. The maximal TEOAE detection improvement is dependent on the coherence function between biological noise in left and right ears. The experimental results show maximal improvement of 7 dB in S/N, improvement in reproducibility near 40% and reduction in duration of TEOAE measurement of over 30%.

Otoacoustic emissions and tympanometry in a general adult population in Sweden: Emisiones otoacústicas y timpanometría en la población general adulta de Suecia

The present study concerns a general adult population in Sweden, not exposed to hazardous occupational noise. Tympanometry and spontaneous (SOAE), transient evoked (TEOAE) and distortion-product (DPOAE) otoacoustic emissions were investigated in 493 randomly selected men and women aged from 20 to 79 years. Effects of gender, age, ear side and middle ear state were determined, with and without adjusting for hearing threshold level. No statistically significant effects on middle ear pressure or compliance were present. For TEOAEs and DPOAEs, the effect of gender and age was statistically significant (p < 0.01), with larger signal levels for female subjects and young subjects, even after adjusting for hearing threshold level. No effect of middle ear pressure on otoacoustic emissions was present, but high middle ear compliance was associated with low emission levels (p < 0.01). Reference data for middle ear compliance and pressure and prevalence data on SOAEs, TEOAEs and DPOAEs for male and female subjects in different age groups were determined. Mean signal levels of TEOAEs and DPOAEs are presented.

Analysis of parameters for the estimation of loudness from tone-burst otoacoustic emissions

There is evidence that tone-burst otoacoustic emissions (TBOAEs) might be useful for estimating loudness. However, within-listener comparisons between loudness and TBOAE measurements are an essential prerequisite to determine appropriate analysis parameters for loudness estimation from TBOAE measurements. The purpose of the present work was to collect TBOAE measurements and loudness estimates across a wide range of levels in the same listeners. Therefore, TBOAEs were recorded for 1- and 4-kHz stimuli and then analyzed using a wide range of parameters to determine which parameter set yielded the lowest mean-square-error estimation of loudness with respect to a psychoacoustical, cross-modality-matching procedure and the inflected exponential (INEX) loudness model. The present results show strong agreement between 1-kHz loudness estimates derived from TBOAEs and loudness estimated using cross-modality matching (CMM), with TBOAE estimation accounting for almost 90% of the CMM variance. Additionally, the results indicate that analysis parameters may vary within a reasonable range without compromising the results (i.e., the estimates exhibit some parametric robustness). The lack of adequate parametric optimization for TBOAEs at 4 kHz suggests that measurements at this frequency are strongly contaminated by ear-canal resonances, meaning that deriving loudness estimates from TBOAEs at this frequency is significantly more challenging than at 1 kHz.

Transient-evoked otoacoustic emissions in a group of professional singers who have normal pure-tone hearing thresholds

OBJECTIVES

The objective of this study was to determine whether transient-evoked otoacoustic emissions (TEOAEs) measured in a group of normal-hearing professional singers, who were frequently exposed to high-level sound during rehearsals and performances, differed from those measured in age- and gender-matched normal-hearing non-singers, who were at minimal risk of hearing loss resulting from excessive sound exposure or other risk factors.

DESIGN

Twenty-three normal-hearing singers (NH-Ss), 23 normal-hearing controls (NH-Cs), and 9 hearing-impaired singers (HI-Ss) were included. Pure-tone audiometry confirmed normal-hearing thresholds (>or=15 dB HL) at 0.5, 1.0, 2.0, 3.0, 4.0, 6.0, and 8.0 kHz in NH-Ss and NH-Cs, and confirmed mild, high frequency, sensorineural hearing loss in HI-Ss (HI-Ss were included only to estimate sensitivity and specificity values for preliminary pass or fail criteria that could be used to help identify NH-Ss at risk for music-induced hearing loss). TEOAEs were measured twice in all ears. TEOAE signal to noise ratio (S/N) and reproducibility were examined for the whole wave response, and for frequency bands centered at 1.0, 1.4, 2.0, 2.8, and 4.0 kHz.

RESULTS

Moderate to high correlations were found between test and retest TEOAE responses for the three groups. However, absolute test-retest differences revealed standard deviations that were two to three times larger than those reported previously, with the majority of the variability occurring for the 1.0 kHz band. As such, only the best TEOAE response (B-TEOAE) from the two measurements in each ear was used in further analyses, with data from the 1.0 kHz band excluded. With one exception, within-group comparisons of B-TEOAE S/N and reproducibility across ears and gender revealed no statistically significant differences for either NH-Ss or NH-Cs. The only significant within-group difference was between left and right ears of NH-C females for S/Ns measured in the 2.0 kHz band, where median responses from right ears were found to be higher than left ears. Across-group comparisons of B-TEOAEs revealed lower median S/N and reproducibility values for NH-Ss compared with NH-Cs for the whole wave response and 1.4 kHz band. For the 2.0 kHz band, reproducibility was similar for the normal-hearing groups but median S/N was found to be lower for NH-Ss. No significant differences in S/N or reproducibility were found between normal-hearing groups for the 2.8 and 4.0 kHz bands. Using data from NH-Cs and HI-Ss to establish sensitivity and specificity values for various TEOAE pass or fail criteria, six preliminary criteria were identified as having sensitivity and specificity values >or=90%. When these criteria were applied to NH-Ss, the number of NH-S ears passing ranged from 57% to 76%, depending on the criteria used to judge the NH-S ears, which translates into 24% to 43% of ears failing.

CONCLUSIONS

Although TEOAE responses were measurable in all singers with normal audiometric thresholds, responses were less robust than those of NH-Cs. The findings suggest that subtle cochlear dysfunction can be detected with TEOAE measurement in a subset of normal-hearing professional singers. Although preliminary, the study findings highlight the importance of pass or fail criterion choice on the number of ears that will be identified as "at risk" for music-induced hearing loss.

Multiresolution analysis of transient evoked otoacoustic emission

To improve transient evoked otoacoustic emission (TEOAE) test accuracy, a new technique in presented based on multiresolution signal decomposition. Using a wavelet function which optimized according to an active cochlear model, time frequency distribution of TEOAE is obtained. The coefficients of the filter bank are determined from Morlet mother wavelet function in the analysis section. In the synthesis section, time windows are obtained from deduced latency curve. By applying this multiresolution analysis, the correlative noise would be canceled. Therefore, the wave reproducibility can be increased without loss of measuring reliability. In abnormal cases where response is not expected the correlative noise is considered as response are completely disappeared. Thus this will increase the accuracy of screening results.

Autoregressive spectrum of transient evoked otoacoustic emissions: methods and applications

The transient evoked otoacoustic emissions (TEOAEs), elicited by a transient sound, is similar to the impulse (or white noise for a random process) response of a system. It fully reveals the information of cochlea in frequency domain. The aim of present study is to apply the autoregressive (AR) spectrum analysis to TEOAEs signal in our OAEs detection system and to find a new quantitative index of cochlear status. The AR spectrum is also applied to TEOAEs with contralateral acoustic stimulation (CAS) to study the function of medial olivocochlear system. The results show clearly that the AR spectrum of TEOAEs signal can be a promising method for diagnosing the impairment of cochlea and medial olivocochlear system.

A Novel Design of a Portable Otoacoustic Emissions Detecting System Based on ARM

Most traditional designs of otoacoustic emissions (OAEs) detecting system adopt the model of "sensor + sound signal sampler / generator + personal computer". This kind of model has shortcomings that it is apt to introduce exterior noises and lacks system integration, which often brings much inconvenience into the process of measuring. There are also some other OAEs detecting systems that use a digital signal processor (DSP) to collect and process the OAEs response signals, but a DSP is not a best choice to perform as a control unit on a portable device for that they usually do not have many peripheral hardware interfaces and are relatively demanding in cost. Different from the fore-mentioned two kinds of control unit, ARM is a kind of elegant new design, with a complete set of common system peripherals, to provide hand-held devices with low-cost and high-performance microcontroller solution in small die size. In this article, a novel design of a highly integrated and network-accessed portable otoacoustic emissions detecting system based on ARM (mainly for TEOAE and DPOAE) is presented. In addition, a comparison of the differences in the system's real-time performance based on different microcontroller units is provided to justify the use of ARM.

Time frequency analysis for transient evoked otoacoustic emissions in different time windows

Transient evoked otoacoustic emissions (TEOAEs) are widely applied in the field of early identification of hearing loss and consequent diagnosis. The duration of testing will not only determine the speed directly but also influence the quality indirectly. In this paper, the default 20ms TEOAEs were divided averagely into 16 continuous 1.25ms-long sub-windows and cross correlation coefficient in each subsection was calculated and compared. Then time-frequency analysis of cone-shaped kernel was applied to demonstrate the frequency changes when TEOAEs were processed under different time windows. It is shown that even if the window was reduced to 12.5 ms, high and middle frequency components were hardly affected while the low frequency component was partially impaired.

The Study About the Reduction of the Stimulus Artifact Using PSR Filter

In order to reduce the stimulus artifact, which is contaminated in the initial segments of the Transient Evoked Otoacoustic Emission(TEOAE) and inhibited the detection of high frequency cochlear responses, an artifact reduction technique that draws on Predictor-Subtracter-Restorer filter for TEOAE signal is firstly proposed. Compared with Derived Nonlinear Response (DNLR) method, the result showed empirically that this proposed method not only reduced the stimulus artifact, but also reserved the useful information of TEOAE at high frequency.

Multiscale Detection of Transient Evoked Otoacoustic Emissions

This paper presents a unified approach to multiscale detection of transient evoked otoacoustic emissions (TEOAEs). Using statistical detection theory, it is shown that the optimal detector involves a time windowing operation where the window can be estimated from ensemble correlation information. The detector performs adaptive splitting of the signal into different frequency bands using either wavelet or wavelet packet decomposition. A simplified detector is proposed in which signal energy is omitted. The results show that the simplified detector performs significantly better than existing TEOAE detectors based on wave reproducibility or the modified variance ratio, whereas the detector involving signal energy does not offer such a performance advantage.

Detection of transient-evoked otoacoustic emissions and the design of time windows

A new approach to the design of time windows is presented for detection of transient-evoked otoacoustic emissions (TEOAE). The windows are designed with reference to a minimum mean square error criterion involving the correlation properties of the ensemble of responses. Latency information is introduced in the detection process by windowing at different scales that result from wavelet decomposition. The significance of both subject- and population-specific time windows is investigated. The detection performance is evaluated on a health screen database consisting of 4989 records. The results show that the present approach to windowing yields a significantly better performance in separating normal-hearing subjects from hearing-impaired subjects when compared to detection based on unwindowed signals. With time windowing, the specificity increased with almost 15% at a fixed sensitivity of 90%.

Estimation of distortion product otoacoustic emissions

Distortion product otoacoustic emissions (DPOAE's) are acoustic signals generated from the cochlea when stimulated by dual tone stimulus. The measurement of the DPOAE's is useful in objectively assessing the hearing functionality of humans. The parameter estimation of the DPOAE's has been based on the fast Fourier transform (FFT) technique, which may not guarantee optimum performance in general. In this paper, we investigate the optimal maximum-likelihood estimator (MLE) for the DPOAE's. Experiments showed that the performance of the MLE is superior to those of the conventional FFT estimators. Furthermore, it is proven that the unwindowed FFT estimator is essentially the MLE when the stimulus tones are constrained to exhibit complete cycle within the data frame.

Synthesis and decomposition of transient-evoked otoacoustic emissions based on an active auditory model

In this paper, transient-evoked otoacoustic emissions (TEOAE's) are synthesized by an active auditory model, and decomposed by continuous wavelet transform (CWT) to study the frequency-latency relationship and the generation of TEOAE signals. The controlled voltage sources that relate to the mobile mechanism of the outer hair cells (OHC's) are proposed to serve as the generation sources of TEOAE signals. The state-variable method is adopted to calculate the auditory model. The mother wavelet used in CWT is selected on basis of the model. The results of this study show that the simulated TEOAE signal is similar to the clinically detected ones not only in the time-domain waveform, but also in the frequency-latency relationship. It seems to be clear that the generation of TEOAE signals is related to the same active mechanism as the cochlear sharp frequency selectivity.

Comparison of the auditory-evoked brainstem response wave I to distortion-product otoacoustic emissions resulting from changes to inner ear blood flow

OBJECTIVE/HYPOTHESIS

Examine and compare in detail the time courses of the auditory brainstem response (ABR) wave I amplitude and latency to the distortion-product otoacoustic emission (DPOAE) amplitude and phase measured in the rabbit model following deliberate obstruction of cochlear blood flow (CBF).

METHODS

Using a posterior fossa craniotomy in five rabbits, the internal auditory artery (IAA) was compressed with a probe. ABR and otoacoustic emission were continuously monitored before, during, and after the compressions.

RESULTS

ABR wave I amplitudes demonstrated measurable decreases at a mean of 28.3 s after IAA compression, whereas DPOAE amplitudes decreased after a mean of 14.8 s. Wave I latencies began to increase at a mean of 18.3 s after occlusion, while DPOAE phase measures changed after a mean of only 4.8 s following IAA compression. The time-course patterns were similar for the amplitudes of both ABR wave I and DPOAE.

CONCLUSIONS

ABR wave I amplitude follows a similar, though delayed (by approximately 10 s) time-course pattern to that of the DPOAE following IAA compression. The implication of these findings for intraoperative auditory monitoring is that changes in many currently employed measures will lag actual surgically induced alterations in CBF by at least 20 to 30 s.

A fast recursive-least-squares adaptive notch filter and its applications to biomedical signals

A fast recursive-least-squares (FRLS) adaptive notch filter (ANF) for cancellation of sinusoidal interference from recorded biomedical signals is investigated. The FRLS ANF is derived by making an approximation to the conventional recursive-least-squares (RLS) ANF for computation economy. It outperforms the commonly adopted least-mean-squares (LMS) ANF, demonstrating a rapid and bandwidth-insensitive initial convergence. A novel application of the FRLS ANF is for the elimination of the tonal artefact in distortion product otoacoustic emission (DPOAE) signals.

On the existence of an age/threshold/frequency interaction in distortion product otoacoustic emissions

Interactions among age, threshold, and frequency in relation to distortion product otoacoustic emissions (DPOAE) have yet to be resolved. The effects of these variables were explored by analyzing DPOAEs in ears with thresholds not exceeding 20 dB HL. Multivariate regression analyses were performed in two different ways. For data to be included in the first analysis, audiometric threshold had to be 20 dB HL or better only at the particular frequency under study, but might exceed 20 dB HL at other half-octave frequencies. Significant main effects were found for age, threshold, and frequency. There was also an age-by-frequency interaction, but a significant age-by-threshold interaction was not observed. DPOAE amplitudes decreased as either age, frequency, or threshold increased. In the second analysis, when a more stringent inclusion criterion was applied (normal thresholds at all frequencies), the main effects for age, threshold, and frequency were not significant. The significant age-by-frequency interaction remained, whereby DPOAE amplitudes decreased as age and frequency increased, but the age-by-threshold interaction again was not significant. The magnitude of DPOAE amplitude change across age, threshold, and frequency and for the age-by-frequency interaction was small but similar for both groups of subjects. Age in association with threshold did not account for observed changes in DPOAE amplitudes for either group. Importantly, the lack of a significant age-by-threshold interaction indicates that there may be processes intrinsic to aging alone that act on DPOAE generation.

Measurement of distortion product phase in the ear canal of the cat

Amplitudes of odd order distortion products (DPs) that are detected in animal ear canals have been used to probe cochlear health, to search for cochlear amplification, and to measure aspects of cochlear mechanical frequency response. Like the DP amplitude, DP phase is also an important measure of the cochlear mechanical response. Reported here are measurements of DP phase in the ear canal of the cat. The phase data show frequency-dependent time delays. One of these delays is a function of f2, the frequency of the higher-frequency primary. Hence the DP phase phi d is of the form phi d = phi 0 + omega d tau, where omega d is the DP angular frequency and tau is a fixed time delay. Our results show that phi d is independent of input level a2 as long as the ratio a2/a1 < or = 2, where a2 and a1 are the amplitudes of the input tones. As a2/a1 becomes greater than two, the fixed time delays increase for DPs whose frequencies are less than the frequencies of the input tones. When both levels are varied together the delay increases as the levels decrease. There can be phase changes as large as pi associated with deep nulls in the DP magnitude for the two lower-frequency DPs. Features of the nulls may be modeled assuming that there is partial reflection of the DP wave from the DP place. The assumption of energy remitted from the DP place also explains amplitude-ratio-dependent time delays and 2 pi level-dependent bifurcations in phase. The DP phase shows different dependencies for f2 < 1 kHz compared to f2 > 2 kHz.