Automated auditory response detection: Statistical problems with repeated testing Evaluación repetida en la detección de respuestas auditivas

Feasibility of Diagnosing Dead Regions Using Auditory Steady-State Responses to an Exponentially Amplitude Modulated Tone in Threshold Equalizing Notched Noise, Assessed Using Normal-Hearing Participants

The aim of this study was to assess feasibility of using electrophysiological auditory steady-state response (ASSR) masking for detecting dead regions (DRs). Fifteen normally hearing adults were tested using behavioral and electrophysiological tasks. In the electrophysiological task, ASSRs were recorded to a 2 kHz exponentially amplitude-modulated tone (AM2) presented within a notched threshold equalizing noise (TEN) whose center frequency (CFNOTCH) varied. We hypothesized that, in the absence of DRs, ASSR amplitudes would be largest for CFNOTCH at/or near the signal frequency. In the presence of a DR at the signal frequency, the largest ASSR amplitude would occur at a frequency (fmax) far away from the signal frequency. The AM2 and the TEN were presented at 60 and 75 dB SPL, respectively. In the behavioral task, for the same maskers as above, the masker level at which an AM and a pure tone could just be distinguished, denoted AM2ML, was determined, for low (10 dB above absolute AM2 threshold) and high (60 dB SPL) signal levels. We also hypothesized that the value of fmax would be similar for both techniques. The ASSR fmax values obtained from grand average ASSR amplitudes, but not from individual amplitudes, were consistent with our hypotheses. The agreement between the behavioral fmax and ASSR fmax was poor. The within-session ASSR-amplitude repeatability was good for AM2 alone, but poor for AM2 in notched TEN. The ASSR-amplitude variability between and within participants seems to be a major roadblock to developing our approach into an effective DR detection method.

Automatic audiometry using auditory steady-state response and sequential test strategy applied to volunteers with normal hearing

PURPOSE

In the present study, a new procedure to perform automatic audiometry using multifrequency Auditory Steady-State Response (ASSR) is proposed.

METHODS

The automatic audiometry procedure consists of detecting the presence of multifrequency ASSR in real-time using the sequential test strategy and by adjusting the stimulus intensity independently. The ASSR audiometric thresholds of 18 adult volunteers with normal hearing were determined by automatically (four simultaneous frequencies per ear) at modulation frequencies in the 80 Hz range. The exam time and the difference between ASSR thresholds and pure-tone behavioural hearing thresholds were estimated as performance measures.

RESULTS

The results showed that automatic audiometry can reduce the number of intensity levels used to obtain the ASSR threshold by up to 58% when compared to audiometry without using the techniques applied in automatic audiometry. In addition, the average of the difference between ASSR thresholds and Pure-Tone Behavioural Hearing thresholds was around 19 dB, which is similar to the results reported in similar studies.

CONCLUSIONS

The audiometric procedure proposed in this study is fully automatic, i.e., does not require any human supervision throughout the exam, and is able to significantly reduce the conventional exam time.

Automated cortical auditory response detection strategy

Objective: This study describes a new automated strategy to determine the detection status of an electrophysiological response.Design: Response, noise and signal-to-noise ratio of the cortical auditory evoked potential (CAEP) were characterised. Detection rules were defined: when to start testing, when to conduct subsequent statistical tests using residual noise as an objective criterion, and when to stop testing.Study sample: Simulations were run to determine optimal parameters on a large combined CAEP data set collected in 45 normal-hearing adults and 17 adults with hearing loss.Results: The proposed strategy to detect CAEPs is fully automated. The first statistical test is conducted when the residual noise level is equal to or smaller than 5.1 µV. The succeeding Hotelling's T² statistical tests are conducted using pre-defined residual noise levels criteria ranging from 5.1 to 1.2 µV. A rule was introduced allowing to stop testing before the maximum number of recorded epochs is reached, depending on a minimum p-value criterion.Conclusion: The proposed framework can be applied to systems which involves detection of electrophysiological responses in biological systems containing background noise. The proposed detection algorithm which optimise sensitivity, specificity, and recording time has the potential to be used in clinical setting.

Faster automatic ASSR detection using sequential tests

Objective: Objective Response Detection (ORD) can be used for auditory steady-state response (ASSR) detection. In conventional ORD methods, the statistical tests are applied at the end of data collection ('single-shot tests'). In sequential ORD methods, statistical tests are applied repeatedly, while data is being collected. However, repeated testing can increase False Positive (FP) rates. One solution is to infer that response is present only after the test remains significant for a predefined number of consecutive detections (NCD). Thus, this paper describes a new method for finding the required NCD that control the FP rate for ASSR detection.Design: NCD values are estimated using Monte Carlo simulations.Study sample: ASSR signals were recorded from 8 normal-hearing subjects.Results: The exam time was reduced by up to 38.9% compared to the single-shot test with loss of approximately 5% in detection rate. Alternatively, lower gains in time were achieved for a smaller (non-significant) loss in detection rate. The FP rates at the end of the test were kept at the nominal level expected (1%).Conclusion: The sequential test strategy with NCD as the stopping criterion can improve the speed of ASSR detection and prevent higher than expected FP rates.

A group sequential test for ABR detection

Objective: To detect the auditory brainstem response (ABR) automatically using an innovative sequentially applied Hotelling's T ² test, with the overall goal of optimising test time whilst controlling the false-positive rate (FPR). Design: The stage-wise critical decision boundaries for accepting or rejecting the null hypothesis were found using a new approach called the Convolutional Group Sequential Test (CGST). Specificity, sensitivity, and test time were evaluated using simulations and subject recorded data. Study sample: Data consists of click-evoked ABR threshold series from 12 normal hearing adults, and recordings of EEG background activity from 17 normal hearing adults. Results: Reductions in mean test time of up to 40-45% were observed for the sequential test, relative to a conventional "single shot" test where the statistical test is applied to the data just once. To obtain these results, it will occasionally be necessary to run the test to a higher number of stimuli, i.e. the maximum test time needs to be increased. Conclusions: The CGST can be used to control the specificity of a sequentially applied ABR detection method. Doing so can reduce test time, relative to the "single shot" test, when considered across a cohort of test subjects.

The Convolutional Group Sequential Test: Reducing Test Time for Evoked Potentials

When using a statistical test for automatically detecting evoked potentials, the number of stimuli presented to the subject (the sample size for the statistical test) should be specified at the outset. For evoked response detection, this may be inefficient, i.e., because the signal-to-noise ratio (SNR) of the response is not known in advance, the user would usually err on the cautious side and use a relatively high number of stimuli to ensure adequate statistical power. A more efficient approach is to apply the statistical test repeatedly to the accumulating data over time, as this allows the test to be stopped early for the high SNR responses (thus reducing test time), or later for the low SNR responses. The caveat is that the critical decision boundaries for rejecting the null hypothesis need to be adjusted if the intended type-I error rate is to be obtained. This study presents an intuitive and flexible method for controlling the type-I error rate for sequentially applied statistical tests. The method is built around the discrete convolution of truncated probability density functions, which allows the null distribution for the test statistic to be constructed at each stage of the sequential analysis. Because the null distribution remains tractable, the procedure for finding the stage-wise critical decision boundaries is greatly simplified. The method also permits data-driven adaptations (using data from previous stages) to both the sample size and the statistical test, which offers new opportunities to speed up testing for evoked response detection.

Automated detection of auditory response: applying sequential detection strategies with constant significance level to magnitude-squared coherence

Abstracts Objective: The detection of the auditory steady-state responses is usually performed by an appropriate objective response detector applied to stimulus-related epochs of the raw electroencephalogram (EEG). In order to improve the detection time, sequential detection strategies are usually used. These multiple tests strategies increase the probability of mistakenly detecting a response. The aim of this study was to develop strategies to determine the critical values for the sequential detection strategies based on constant significance level tests. Design: Extensive Monte Carlo simulations were used to test these strategies for the magnitude-squared coherence (MSC) detector. The performances of these strategies were compared with previous works found in the literature. Study sample: All strategies were applied to synthetic and real EEG datasets. Results: The strategies ensure the desired significance level at the end of the sequential detection strategy. The simulated results are in accordance with the real data results. Conclusions: For the MSC detector, where the critical value depends on the number of epochs, the proposed sequential detection strategies obtain better performance regarding test time and detection rate, but worse overall detection rate compared to applying a test only once.

Evaluation of Speed and Accuracy of Next-Generation Auditory Steady State Response and Auditory Brainstem Response Audiometry in Children With Normal Hearing and Hearing Loss

OBJECTIVES

The first objective of this study was to compare the predicted audiometric thresholds obtained by auditory steady state response (ASSR) and auditory brainstem response (ABR) in infants and toddlers when both techniques use optimal stimuli and detection algorithms. This information will aid in determining the basis for large discrepancies in ABR and ASSR measures found in past studies. The hypothesis was that advancements in ASSR response detection would improve (lower) thresholds and decrease discrepancies between the thresholds produced by the two techniques. The second objective was to determine and compare test times required by the two techniques to predict thresholds for both ears at the 4 basic audiometric frequencies of 500, 1000, 2000, and 4000 Hz.

DESIGN

A multicenter clinical study was implemented at three university-based children's hospital audiology departments. Participants were 102 infants and toddlers referred to the centers for electrophysiologic testing for audiometric purposes. The test battery included wideband tympanometry, distortion-product otoacoustic emissions, and threshold measurements at four frequencies in both ears using ABR and ASSR (randomized) as implemented on the Interacoustics Eclipse systems with "Next-Generation" ASSR detection and FMP analysis for ABR. Both methods utilized narrow band CE-Chirp stimuli. Testers were trained on a specialized test battery designed to minimize test time for both techniques. Testing with both techniques was performed in one session. Thresholds were evaluated and confirmed by the first author and correction factors were applied. Test times were documented in system software.

RESULTS

Corrected thresholds for ABR and ASSR were compared by regression, by the Bland-Altman technique and by matched pairs t tests. Thresholds were significantly lower for ASSR than ABR. The ABR-ASSR discrepancy at 500 Hz was 14.39 dB, at 1000 Hz was 10.12 dB, at 2000 Hz was 3.73 dB, and at 4000 Hz was 3.67 dB. The average test time for ASSR of 19.93 min (for 8 thresholds) was found to be significantly lower (p < 0.001) than the ABR test time of 32.15 min. One half of the subjects were found to have normal hearing. ASSR thresholds plotted in dB nHL for normal-hearing children in this study were found to be the lowest yet described except for one study which used the same technology.

CONCLUSIONS

This study found a reversal of previous findings with up to 14 dB lower thresholds found when using the ASSR technique with "Next-Generation" detection as compared with ABR using an automated detection (FMP). The test time for an audiogram prediction was significantly lower when using ASSR than ABR but was excellent by clinical standards for both techniques. ASSRs improved threshold performance was attributed to advancements in response detection including utilization of information at multiple harmonics of the modulation frequency. The stimulation paradigm which utilized narrow band CE-Chirps also contributed to the low absolute levels of the thresholds in nHL found with both techniques.

Automated auditory response detection: Further improvement of the statistical test strategy by using progressive test steps of iteration

OBJECTIVE

A common testing approach for automated ABR detection is to use a sequential test strategy. Repeated testing increases the error probability for a falsely detected response (Type I error rate). To compensate for this effect, the statistically critical test value must be increased with each test step. The aim of the study was to improve response detection by a reduction of the number of test steps using a progressively increasing test step of iteration, defined here as step width.

DESIGN

A progressively increasing test step width was tested with and without the table-related testing (adjusting the critical test value to each test step) proposed by Stürzebecher & Cebulla (2013) . For this study the same data pool was used.

STUDY SAMPLE

The investigation was performed on raw EEG data collected during routine clinical measurement of frequency-specific ASSR for hearing threshold assessment.

RESULTS

The reduction of the test step number combined with a progressive test step width led to a significantly improved response detection. In combination with table-related testing a slight but not significant improvement compared to table-related testing alone was revealed.

CONCLUSIONS

The proposed test strategy can improve the performance of objective hearing threshold assessment and of newborn hearing screening.

Habituation of Auditory Steady State Responses Evoked by Amplitude-Modulated Acoustic Signals in Rats

Generation of the auditory steady state responses (ASSR) is commonly explained by the linear combination of random background noise activity and the stationary response. Based on this model, the decrease of amplitude that occurs over the sequential averaging of epochs of the raw data has been exclusively linked to the cancelation of noise. Nevertheless, this behavior might also reflect the non-stationary response of the ASSR generators. We tested this hypothesis by characterizing the ASSR time course in rats with different auditory maturational stages. ASSR were evoked by 8-kHz tones of different supra-threshold intensities, modulated in amplitude at 115 Hz. Results show that the ASSR amplitude habituated to the sustained stimulation and that dishabituation occurred when deviant stimuli were presented. ASSR habituation increased as animals became adults, suggesting that the ability to filter acoustic stimuli with no-relevant temporal information increased with age. Results are discussed in terms of the current model of the ASSR generation and analysis procedures. They might have implications for audiometric tests designed to assess hearing in subjects who cannot provide reliable results in the psychophysical trials.

Automated auditory response detection: Improvement of the statistical test strategy

OBJECTIVE

Automated auditory response detection is always performed by applying an appropriate statistical test to a sample of stimulus-related epochs of the raw EEG. The often-used sequential test strategy saves time, but the multiple testing increases the probability of falsely detected responses. Therefore, the critical test value must guarantee the specified error probability for the maximum test step number. However, response detection at all lower test step numbers is disadvantaged. We propose calculating the critical test values for each test step number, which correspond exactly to the given error probability.

DESIGN

The critical test values for each test step were calculated by the method described by Stürzebecher et al (2005) . A table with the test values was implemented with customized software of the Eclipse ASSR system(®) (Interacoustics, Denmark).

STUDY SAMPLE

Table-related testing was performed on a sample of raw EEG data collected during the routine clinical measurement of frequency-specific auditory steady-state responses (ASSR) for hearing threshold assessment.

RESULTS

The new test strategy leads to a significantly increased detection rate and a significantly shorter detection time.

CONCLUSIONS

The new test strategy can improve the performance of the objective hearing threshold assessment and of the newborn hearing screening.

Establishing auditory steady-state response thresholds to narrow band CE-chirps(®) in full-term neonates

Narrow band CE-chirps(®) were developed to provide a better synchronization of neural response due to the compensation of the traveling wave delay in the basilar membrane. These stimuli combined with a detection method that includes higher response harmonics on the auditory steady-state response (ASSR) recording was studied in this research.

OBJECTIVE

(1) To establish air conduction thresholds for ASSR to narrow band CE-chirp(®) in normal hearing full-term neonates; (2) describe the test time needed for the above in one ear and (3) to compare the results in infants and normally hearing adults.

METHOD

ASSR to air-conducted stimuli were obtained in 30 full-term neonates (14 girls and 16 boys) with an average age of 34.3h of life. All neonates were presented presence of transient-evoked otoacoustic emissions (TEOAE) and result "pass" in automatic ABR at 35dB nHL before ASSR test. ASSR thresholds of both ears of 10 normal hearing adults (5 girls and 5 boys) varied in age between 23 and 30 years and with hearing thresholds better than or equal to 15dB HL at all frequencies between 250 and 8000Hz were recorded to compare with the neonate data.

RESULTS

The neonate ASSR thresholds estimated from 50% using cumulative distributions were 24.5, 13.5, 7.5 and 10dB nHL at 500, 1000, 2000, and 4000Hz, respectively. For the same frequency order, ASSR thresholds estimated from 90% of the neonates were 34.5, 28, 12.5 and 15dB nHL. It required 21.2 (±5)min on average to obtain threshold in each ear in neonates, with a range of 12-29min. When ASSR thresholds recorded in full-term neonates and adults were directly compared, the differences between these groups were not significant for 1000Hz (p=0.500), 2000Hz (p=0.610) and 4000Hz (p=0.362). However, at 500Hz, ASSR thresholds in neonates tend to be greater than in adults (p=0.001).

CONCLUSION

In this study ASSR thresholds estimated from 90% of the neonates were 34.5, 28, 12.5 and 15dB nHL. It required 21.2 (±5)min on average to obtain threshold in each ear and ASSR thresholds to narrow band CE-chirp(®) in neonates are not significant for adults ASSR thresholds, except at 500Hz, when the ASSR thresholds in neonates tend to be greater than in adults.

Auditory steady state responses in normal-hearing and hearing-impaired adults: an analysis of between-session amplitude and latency repeatability, test time, and F ratio detection paradigms

OBJECTIVES

The aim of this study was to assess the between-session repeatability of auditory steady state response (ASSR) amplitudes and to examine F ratio response detection parameters.

DESIGN

Suprathreshold ASSRs were recorded from 20 normal-hearing and 10 hearing-impaired subjects. Amplitudes and latencies were recorded in two test sessions conducted on separate days.

ANALYSIS

The repeatability coefficients (limits of expected variation in repeat measurements) for amplitude and latency of ASSRs were calculated. The test time required for the responses to reach significance at 1%, 2%, and 5% F ratios was analyzed. The percentage false response detection rate was calculated to determine the suitability of current ASSR threshold estimation protocols for use in audiology clinics.

RESULTS

The repeatability coefficients for the amplitude of ASSRs were 29 nV for the normal-hearing subjects and 57 nV for the hearing-impaired subjects. The repeatability coefficients for the latency of ASSR were 1.10 msec for the normal-hearing subjects and 1.19 msec for the hearing-impaired subjects. High false-positive detection rates were found for detection procedures that used variable test time ("stop when significance reached" methods).

CONCLUSIONS

The results of this study showed that ASSR amplitudes are highly variable between test sessions with an average estimated variability in response amplitude of ± 40% for normal-hearing participants and ± 97% for hearing-impaired participants. This could be a possible cause of test-retest differences in ASSR threshold measurements, as it could potentially lead to thresholds that were above the EEG noise level and significant in one test session subsequently falling below the EEG noise level in the repeat test session leading to insignificant response and thus poorer ASSR threshold.

Roots of Adult Attachment

It is widely assumed that, within the context of a stable developmental environment, relationship experiences in early life influence later ones. To date, however, there has been no longitudinal empirical evidence for the hypothesis that early maternal caregiving predicts adult attachment dynamics with peers and partners. The present longitudinal study shows that quality of maternal caregiving experienced at 18 months of age predicted the extent to which the same participants more than 20 years later (age M = 22) were uncomfortable relying on partners and peers (avoidance) and experienced relational worries with partners (anxiety). These findings provide new empirical support that early maternal caregiving predicts later adult attachment patterns with peers and partners. Moreover, consistent with attachment theory, they suggest that the influence of maternal caregiving experienced in early life is not limited to this first attachment relationship but operates more generally in other attachment relationships.

Monitoring residual hearing during cochlear implantation by intra-operative brainstem audiometry

OBJECTIVE

Aim of this paper is to prove the applicability of intra-operative recordings of auditory brainstem responses during cochlear implantation.

METHODS

The clinical practicability of intra-operative monitoring of hearing thresholds (Notched-Noise BERA, Amplitude Modulation Following Response [AMFR]) is presented in the respective case. The recordings were performed prior to the cochlear implantation and were compared with those obtained during and after cochlear implantation.

RESULTS

It is demonstrated that the patient's cochlear function can be monitored; residual hearing is available after surgery.

CONCLUSION

The possibility of monitoring of hearing thresholds may add some security to the concept of electric-acoustic stimulation.

Auditory steady-state responses to chirp stimuli based on cochlear traveling wave delay

This study investigates the use of chirp stimuli to compensate for the cochlear traveling wave delay. The temporal dispersion in the cochlea is given by the traveling time, which in this study is estimated from latency-frequency functions obtained from (1) a cochlear model, (2) tone-burst auditory brain stem response (ABR) latencies, (3) and narrow-band ABR latencies. These latency-frequency functions are assumed to reflect the group delay of a linear system that modifies the phase spectrum of the applied stimulus. On the basis of this assumption, three chirps are constructed and evaluated in 49 normal-hearing subjects. The auditory steady-state responses to these chirps and to a click stimulus are compared at two levels of stimulation (30 and 50 dB nHL) and a rate of 90s. The chirps give shorter detection time and higher signal-to-noise ratio than the click. The shorter detection time obtained by the chirps is equivalent to an increase in stimulus level of 20 dB or more. The results indicate that a chirp is a more efficient stimulus than a click for the recording of early auditory evoked responses in normal-hearing adults using transient sounds at a high rate of stimulation.

Improving Auditory Steady-State Response Detection Using Independent Component Analysis on Multichannel EEG Data

Over the last decade, the detection of auditory steady-state responses (ASSR) has been developed for reliable hearing threshold estimation at audiometric frequencies. Unfortunately, the duration of ASSR measurement can be long, which is unpractical for wide scale clinical application. In this paper, we propose independent component analysis (ICA) as a tool to improve the ASSR detection in recorded single-channel as well as multichannel electroencephalogram (EEG) data. We conclude that ICA is able to reduce measurement duration significantly. For a multichannel implementation, near-optimal performance is obtained with five-channel recordings.

Reliability of electric response audiometry using 80 Hz auditory steady-state responses

The reliability of the Auditory Steady State Response (ASSR) has not been thoroughly evaluated despite its recent application as a clinical tool for threshold estimation. The purpose of this study was to examine test-retest (TR) reliability of ASSR threshold estimates in an empirical research design. The ASSR, tested using modulation frequencies approximately 80 Hz and above, was evaluated against pure tone audiometry (PTA), and the slow vertex potential (SVP, N1-P2). Sixteen normal-hearing young female adults were tested twice, one week apart. Varying degrees of sensorineural hearing loss of a notched configuration were simulated with filtered masking noise. Test-retest reliability was assessed using Pearson-product moment correlation analysis, supplemented by other post-hoc analyses. Results demonstrated moderately strong TR reliability for ASSR at 1000, 2000 and 4000 Hz (r = 0.83-0.93); however, the reliability of ASSR at 500 Hz was weaker (r = 0.75). Results suggest that ASSR-ERA is a reliable test at mid-high frequencies, at least with the configuration and degrees of simulated sensorineural hearing loss examined in this study.