Normal-hearing Adults Research Articles

Early Loss of Spiral Ganglion Neurons in the Auditory System after Noise Trauma

Introduction: Noise-induced hearing loss is one of the most frequent recognized occupational diseases. The time course of the involved pathologies is still under investigation. Several studies have demonstrated an acute damage of the sensory tissue, but only few experiments investigated the degeneration of (type I) spiral ganglion neurons (SGNs), representing the primary neurons in the auditory system. The aim of the present study was to investigate the time course of SGN degeneration within a 7-day period after traumatic noise exposure starting immediately after trauma. Methods: Young adult normal hearing mice were noise exposed for 3 h with a broadband noise (5–20 kHz) at 115 dB SPL. Auditory threshold shift was measured by auditory brainstem recordings, and SGN densities were analyzed at different time points during the first week after acoustic trauma. Results: Significant reduction of SGN densities was detected and is accompanied by a significant hearing loss. Degeneration starts within hours after the applied trauma, further progressing within days post-exposure. Discussion: Early neurodegeneration in the auditory periphery seems to be induced by direct overstimulation of the auditory nerve fibers. SGN loss is supposed to be a result of inflammatory responses and neural deprivation, leading to permanent hearing loss and auditory processing deficits.

Examining Force Level Output of Skin-Drive Bone Conduction Hearing Devices in Adults With Simulated Conductive Hearing Loss.

Bone conduction hearing devices (BCDs) that deliver sound across the skin (i.e., transcutaneous) are suitable for some individuals who have conductive or mixed hearing losses. Prescriptive targets for percutaneous devices are available, for example, from the Desired Sensation Level-Bone Conduction Hearing Device (DSL-BCD) algorithm. These targets, however, may require modification for use with transcutaneous BCDs. The current study investigated three key variables that may inform target modification: (a) comparison of thresholds measured using an audiometric bone conduction (BC) transducer versus transcutaneous BCDs that offer in situ threshold measurement, (b) transcutaneous BCD default force level outputs versus recommended DSL percutaneous BC targets, and (c) the preferred listening levels (PLLs) of adults wearing transcutaneous BCDs in a laboratory setting. Bilateral conductive hearing loss was simulated in 20 normal-hearing adults via earplugs. Thresholds were measured using a B-71 BC transducer and two commercially available BCDs coupled to a soft headband. DSL percutaneous BC targets were generated, and PLLs were obtained for a 60-dB SPL speech stimulus. Force level outputs were measured using a skull simulator on the Audioscan Verifit2 at the hearing aids' default settings and at the participants' PLL for each device. On average, audiometric BC thresholds were significantly better than those measured in situ with each BCD. PLLs were similar to prescribed targets for one device with the smoother response shape and agreed in the high frequencies for both devices. In situ thresholds are significantly higher than audiometric BC thresholds, suggesting that device-based in situ measurement more accurately accounts for the signal transmission from transcutaneous BCDs. PLLs differed from the percutaneous targets and varied between devices, which may indicate that either target modifications or manipulations of device frequency response shaping are needed to approximate PLL with transcutaneous BCD devices.

Neural processing of speech comprehension in noise predicts individual age using fNIRS-based brain-behavior models.

Speech comprehension in noise depends on complex interactions between peripheral sensory and central cognitive systems. Despite having normal peripheral hearing, older adults show difficulties in speech comprehension. It remains unclear whether the brain's neural responses could indicate aging. The current study examined whether individual brain activation during speech perception in different listening environments could predict age. We applied functional near-infrared spectroscopy to 93 normal-hearing human adults (20 to 70years old) during a sentence listening task, which contained a quiet condition and 4 different signal-to-noise ratios (SNR = 10, 5, 0, -5dB) noisy conditions. A data-driven approach, the region-based brain-age predictive modeling was adopted. We observed a significant behavioral decrease with age under the 4 noisy conditions, but not under the quiet condition. Brain activations in SNR = 10dB listening condition could successfully predict individual's age. Moreover, we found that the bilateral visual sensory cortex, left dorsal speech pathway, left cerebellum, right temporal-parietal junction area, right homolog Wernicke's area, and right middle temporal gyrus contributed most to prediction performance. These results demonstrate that the activations of regions about sensory-motor mapping of sound, especially in noisy conditions, could be sensitive measures for age prediction than external behavior measures.

Influence of Talker and Accent Variability on Rapid Adaptation and Generalization to Non-Native Accented Speech in Younger and Older Adults

ABSTRACT Listeners can rapidly adapt to English speech produced by non-native speakers of English with unfamiliar accents. Prior work has shown that the type and number of talkers contained within a stimulus set may impact rate and magnitude of learning, as well as any generalization of learning. However, findings across the literature have been inconsistent, with relatively little study of these effects in populations of older listeners. In this study, adaptation and generalization to unfamiliar talkers with familiar and unfamiliar accents are studied in younger normal-hearing adults and older adults with and without hearing loss. Rate and magnitude of adaptation are modeled using both generalized linear mixed-effects regression and generalized additive mixed-effects modeling. Rate and magnitude of adaptation were not impacted by increasing the number of talkers and/or varying the consistency of non-native English accents across talkers. Increasing the number of talkers did strengthen generalization of learning for a talker with a familiar non-native accent, but not for an unfamiliar accent. Aging alone did not diminish adaptation or generalization. These findings support prior evidence of a limited benefit for talker variability in facilitating generalization of learning for non-native accented speech, and extend the findings to older adults.

Horizontal sound localization in presence of noise in normal-hearing young adults

Objective: This study investigates the effect of signal-to-noise ratio (SNR), frequency, and bandwidth on horizontal sound localization accuracy in normal-hearing young adults. Methods: From August 2022 to December 2022, a total of 20 normal-hearing young adults, including 7 males and 13 females, with an age range of 20 to 35 years and a mean age of 25.4 years, were selected to participate in horizontal azimuth recognition tests under both quiet and noisy conditions. Six narrowband filtered noise stimuli were used with central frequencies (CF) of 250, 2 000, and 4 000 Hz and bandwidths of 1/6 and 1 octave. Continuous broadband white noise was used as the background masker, and the signal-to-noise ratio (SNR) was 0, -3, and -12 dB. The root-mean-square error (RMS error) was used to measure sound localization accuracy, with smaller values indicating higher accuracy. Friedman test was used to compare the effects of SNR and CF on sound localization accuracy, and Wilcoxon signed-rank test was used to compare the impact of the two bandwidths on sound localization accuracy in noise. Results: In a quiet environment, the RMS error in horizontal azimuth in normal-hearing young adults ranged from 4.3 to 8.1 degrees. Sound localization accuracy decreased with decreasing SNR: at 0 dB SNR (range: 5.3-12.9 degrees), the difference from the quiet condition was not significant (P>0.05); however, at -3 dB (range: 7.3-16.8 degrees) and -12 dB SNR (range: 9.4-41.2 degrees), sound localization accuracy significantly decreased compared to the quiet condition (all P<0.01). Under noisy conditions, there were differences in sound localization accuracy among stimuli with different frequencies and bandwidths, with higher frequencies performing the worst, followed by middle frequencies, and lower frequencies performing the best, with significant differences (all P<0.01). Sound localization accuracy for 1/6 octave stimuli was more susceptible to noise interference than 1 octave stimuli (all P<0.01). Conclusions: The ability of normal-hearing young adults to localize sound in the horizontal plane in the presence of noise is influenced by SNR, CF, and bandwidth. Noise with SNRs of ≥-3 dB can lead to decreased accuracy in narrowband sound localization. Higher CF signals and narrower bandwidths are more susceptible to noise interference.

Spectral resolution and voice cue weighting in adults with normal hearing

Cochlear implant (CI) users use acoustic voice cues differently than normal hearing (NH) adults to identify a talker’s gender. Specifically, whereas NH listeners weight a talker’s fundamental frequency (f0) and resonance (operationalized as vocal tract length or VTL) equally, CI users rely almost exclusively on f0. CI users’ abnormal cue weighting may partially arise due to degraded auditory information delivered by the CI. We hypothesized that altering the amount of spectral information in the signal impacts voice cue weighting in NH listeners. Thirty NH adults performed a gender identification task. Auditory stimuli were monosyllabic words synthesized to have one of five f0 values and one of five VTL values. Synthesized voices were processed using a 16-, 8-, and 4-channel noise-band vocoder. Perceptual weights for each voice cue were estimated as the coefficients for f0 and VTL in regression models, with higher coefficients corresponding to stronger perceptual weightings. Listeners relied more on f0 in conditions with less spectral resolution than in conditions with greater spectral resolution. Cue weights for VTL did not change across conditions. Results suggest that removing frequency information from the auditory signal can modify the extent to which listeners use f0 to identify the gender of a talker.

Fundamental dimensions of real-time spoken word recognition in cochlear implant users

Spoken word recognition is a sophisticated cognitive process that maps incoming speech to meaning. For young, normal-hearing adults, word recognition is served by a competition process which plays out incrementally as speech unfolds. Candidates that match the input are activated and compete as mismatching candidates are suppressed. It remains unclear how this process changes in other listeners. Previous results from small-scale studies suggest three dimensions: 1) the speed of activating words (Activation Rate), affected by development and aging; 2) the degree of ultimate competitor suppression (Sustained Activation), affected by hearing loss; and 3) the delay of competition entirely (Wait-and-See), affected by severe hearing loss. To investigate these dimensions, a large, heterogenous group of cochlear implant users (N = 101) completed a Visual World Paradigm task. A principal component analysis supported these three dimensions. Each dimension was predicted by different demographic or auditory factors (onset of deafness, auditory fidelity, age), and each predicted outcomes over and above auditory fidelity. This suggests that these are orthogonal dimensions along which listeners vary, not all-or-nothing strategies. This work identifies the degrees of freedom that can extend theories of word recognition to account for a range of experiences and contexts.

Perceptual consequences of reverberant environments on spatial unmasking

Spatial hearing provides access to auditory spatial cues that promote speech perception in noisy listening situations. However, reverberation degrades auditory spatial cues and limits listeners’ ability to utilize these cues for segregating target speech from competing babble. Hence, spatial unmasking—an intelligibility benefit from a spatial separation between a target and masker—is reduced in reverberant environments as compared to free field. To understand the perceptual consequences of poorer spatial unmasking in reverberation, we assessed three aspects of functional spatial hearing in virtual reverberant environments: perceived auditory source width, auditory spatial acuity, and spatial unmasking. Three auditory environments were simulated and auralized using ODEON to vary interaural coherence (IC): (1) a control anechoic environment, (2) a classroom designed to meet classroom acoustics standards (IC = 0.58), and (3) a classroom of the same size with more severe reverberation (IC = 0.37). Individually measured head-related transfer functions were used to binaurally reproduce the auralized signals over headphones. We hypothesize that interaural decorrelation, the result of increasing reverberation, will broaden the perceived auditory source width with a cascading effect of reduced auditory spatial acuity and subsequently poorer spatial unmasking. Preliminary data from normal-hearing adults will be presented. [Work supported by NIH R21DC020532.]

Perceptual benefits of implicit auditory learning of rule-based nonspeech sound patterns in younger and older adults

Sounds in everyday environments typically precede and follow one another in systematic ways determined by the properties and interactions of corresponding sound sources. Familiar and predictable contextual relationships between consecutive sounds are known to provide perceptual benefit under adverse listening conditions. Previous work has demonstrated implicit learning of arbitrary rule-based sound patterns by young normal hearing adults. However, the role of such rule-based learning on identification of individual sounds within a sound pattern and the effect of listener age has not been investigated. Here younger and older adults with age-appropriate hearing were first familiarized with a set of rule-based nonspeech sound sequences produced by a finite-state machine. Subsequently, they identified individual target sounds within sequences under a range of signal-to-noise ratios (0 to -15 dB). The test sequences were (a) familiar, rule-based; (b) novel, rule-based; or (c) novel, non rule-based. The results for both younger and older listeners revealed superior sound identification in sequences versus in isolation, and a greater overall accuracy of rule-based versus non rule-based sequences. Younger listeners significantly outperformed older listeners. Implicit learning benefit did not correlate with measures of working memory processing. Current findings have implications for audiologic rehabilitation and auditory display design.

Comparison of Two Clinical Devices for the Measurement of Distortion Product Otoacoustic Emissions in Normal-Hearing Adults.

Otoacoustic emissions (OAEs) are low-intensity sounds generated by the cochlea and associated with the function of the outer hair cells. Since OAE measurements do not require active participation of a listener, OAEs are considered an objective measure of cochlear function. While distortion-product OAEs (DPOAEs) are commonly used in clinical practice, limited information is available on the performance of various clinical devices. This study compared two commercial clinical devices, Titan and Eclipse from Interacoustics, and collected normative data for DPOAEs in people with normal hearing. The data collection and analysis were focused on signal and noise amplitudes as well as signal-to-noise ratios (SNRs). Sixty-three participants with normal hearing (age 23.2±2.0 years) were included in the study. DPOAEs were measured at 31 frequencies ranging from 500 to 10,000 Hz. DPOAE amplitude differed across frequencies. Additionally, a high number of unreliable responses were observed at the edge frequencies (<828 Hz and >6,072 Hz). There were no significant differences between the two devices in terms of DPOAE amplitudes, but the recorded-noise levels (amplitude) differed significantly. SNRs of the recorded OAEs were frequency dependent (higher frequencies showed larger SNRs), and significant differences in terms of SNR were found between the two devices. Despite the above-mentioned differences, the OAEs recorded with both devices met the pass criteria for the SNR (≥6 dB) consistently across frequencies, and thus the differences do not compromise the test outcomes. The frequency dependence of the OAE amplitudes and the corresponding SNRs may be relevant for clinical practice.

Evaluating speech-in-speech perception via a humanoid robot.

Underlying mechanisms of speech perception masked by background speakers, a common daily listening condition, are often investigated using various and lengthy psychophysical tests. The presence of a social agent, such as an interactive humanoid NAO robot, may help maintain engagement and attention. However, such robots potentially have limited sound quality or processing speed. As a first step toward the use of NAO in psychophysical testing of speech- in-speech perception, we compared normal-hearing young adults' performance when using the standard computer interface to that when using a NAO robot to introduce the test and present all corresponding stimuli. Target sentences were presented with colour and number keywords in the presence of competing masker speech at varying target-to-masker ratios. Sentences were produced by the same speaker, but voice differences between the target and masker were introduced using speech synthesis methods. To assess test performance, speech intelligibility and data collection duration were compared between the computer and NAO setups. Human-robot interaction was assessed using the Negative Attitude Toward Robot Scale (NARS) and quantification of behavioural cues (backchannels). Speech intelligibility results showed functional similarity between the computer and NAO setups. Data collection durations were longer when using NAO. NARS results showed participants had a relatively positive attitude toward "situations of interactions" with robots prior to the experiment, but otherwise showed neutral attitudes toward the "social influence" of and "emotions in interaction" with robots. The presence of more positive backchannels when using NAO suggest higher engagement with the robot in comparison to the computer. Overall, the study presents the potential of the NAO for presenting speech materials and collecting psychophysical measurements for speech-in-speech perception.

Cortical sensory gating and reactions to dynamic speech-in-noise in older normal-hearing and hearing-impaired adults

Objective To examine whether cortical sensory gating predicts how older adults with and without hearing loss perform the Tracking of Noise Tolerance (TNT) test. Design Single-blind mixed design. TNT performance was defined by average tolerated noise relative to speech levels (TNTAve) and by an average range of noise levels over a two-minute trial (excursion). Sensory gating of P1-N1-P2 components was measured using pairs of 1 kHz tone pips. Study Sample Twenty-three normal-hearing (NH) and 16 hearing-impaired (HI) older adults with a moderate-to-severe degree of sensorineural hearing loss. Results NH listeners tolerated significantly more noise than HI listeners, but the two groups did not differ in their excursion. Both NH and HI listeners exhibited significant gating of P1 amplitudes and N1P2 peak-to-peak amplitudes with no difference in gating magnitudes between listener groups. Sensory gating magnitudes of P1 and N1P2 did not predict TNTAve scores, but N1P2 gating negatively predicted excursion after accounting for listener age and hearing thresholds. Conclusions Listeners’ reactivity to a roving noise (excursion), but not their average noise tolerance (TNTAve), was predicted by sensory gating at N1P2 generators. These results suggest that temporal aspects of speech-in-noise processing may be affected by declines in the central inhibition of older adults.

Feasibility of an Adaptive Version of the Everyday Conversational Sentences in Noise Test.

To investigate potential reasons for the mismatch between laboratory/clinic-based sentence-in-noise performance and real-world listening abilities, we recently developed a corpus of natural, spontaneously spoken speech with three vocal effort levels (Everyday Conversational Sentences in Noise [ECO-SiN]). Here, we examined the feasibility of using the ECO-SiN corpus for adaptive speech-in-noise testing, which might be a desirable format in certain situations (e.g., during a clinical visit). Ten young, normal-hearing adults, along with 20 older adults with hearing loss participated in the study. Speech reception thresholds (SRTs) were obtained using ECO-SiN sentences, which were systematically compared to the SRTs obtained using traditional Bamford-Kowal-Bench-like sentences. Results demonstrated the properties of the test compared favorably with those of a standard test based on scripted and clearly spoken sentences. Moreover, whereas normal-hearing listeners received a benefit from an increase in vocal effort, the participants with hearing loss showed a disbenefit that increased with increasing hearing loss. The adaptive version of the ECO-SiN test is feasible for research and clinical testing. https://doi.org/10.23641/asha.25146338.

Exploring the Effects of Alternate Auditory Attention Tasks on Electromotility of Cochlear Outer Hair Cells in Healthy Normal Hearing Adults

Background and Aim: There is limited study on the role of rostral efferent auditory pathway on Outer Hair Cell (OHC) activity. We investigated the effect of integrating alternate auditory attention (ALAUDIN©) tasks with White Noise (WN) and its reliability using contralateral suppression of Transient Evoked Otoacoustic Emission (TEOAE). Methods: This study was conducted at the Audiology Clinic, Universiti Kebangsaan Malaysia, with fifty normal-hearing adults. All subjects underwent standard audiological testing to ensure normal hearing, middle ear, and cochlear function. Contralateral suppressors with and without attention tasks were delivered randomly to the non-test ear while simultaneously measuring TEOAE amplitude in the test ear to investigate the effect of auditory attention on OHC electromotility. Suppressors with and without attention refer to the combination of a 1000 Hz tone and speech stimulus embedded in WN and WN alone, respectively. Subsequently, the difference in TEOAE amplitude during the presence and absence of suppressors was calculated, and thus suppression magnitude was determined. Results: Intraclass correlation revealed that 4 suppressors produced high reliability. In paired sample t-tests, the tasks significantly reduced the amplitude of the TEOAE in the right ear compared to the left ear (p&lt;0.05). However, the suppression magnitude did not differ significantly between ears (p&gt;0.05). Descriptively, females showed greater TEOAE suppression. Conclusion: This study demonstrated that directing ALAUDIN© tasks to one ear can affect OHC electromotility, as evidenced by TEOAE amplitude changes, but it did not impact the overall suppression magnitude. Additionally, it hinted at potential gender differences in TEOAE suppression, warranting further investigation. Keywords: Suppression; otoacoustic emissions; efferent pathway; auditory attention

Sensory Gating Networks in Normal-Hearing Adults With Minimal Tinnitus.

The goal of this study was to observe sensory gating-related networks underlying cortical auditory evoked potential (CAEP) peak components in individuals with and without minimal tinnitus, as measured using the Tinnitus Handicap Inventory (THI). This analysis was performed on previously published sensory gating responses in normal-hearing adults with and without minimal tinnitus. Independent component analysis was performed for each individual CAEP gating component (Pa, P50, N1, and P2). Significant components were retained for source localization analyses within the following groups: no tinnitus, tinnitus with a THI score ≤ 6, and tinnitus with a THI score > 6. Brain source localization was performed on the gating difference wave for each component using standardized low-resolution brain electromagnetic tomography. Gating-related networks were identified within each group. Different regional sources were observed between groups, with parietal sources underlying the Pa and P50 components as tinnitus severity increased. A larger prefrontal regional activation was also shown for the N1 gating component as tinnitus severity increased. These results expand upon the functional gating responses via CAEP waveforms in a previously published study. The auditory gating response, as measured via CAEPs, has previously been shown to significantly correlate with an increase in tinnitus severity in adults with normal hearing. The corresponding changes in the gating response appear to be supported by different cortical regions in those without tinnitus, those with a THI score ≤ 6, and those with a THI score > 6. Next, functional differences between localized cortical regions should be tested.

Effect of hearing loss on P300 measures

Background: P300 is among the first auditory responses in a collection of event-related or endogenous evoked responses. Hearing loss has drastically reduced cognitive abilities in individual adults. P300 is used as an electrophysiological tool to assess cognitive functions. This study was framed to compare the effect of hearing loss in P300 amplitude and latencies in Indian population. As no such study was carried out in Indian population, this study will be helpful in evaluating the correlation between hearing loss and cognitive function. Methods: We recruited a total of 60 participants out of which 30 were in group 1 having normal hearing and other 30 were included in group 2 having hearing loss. P300 was obtained using a Neurosoft instrument using tone burst stimuli The acquisition of tonal stimuli was carried out in a sound-treated room. Results: The t test value (t=4.75, p&lt;0.001) for P300 amplitude indicates that the difference between the mean among normal hearing and hearing-impaired adult was statistically significant. But the p=0.68 for P300 latency which indicates that the difference between the two groups is not statistically significant, which accepts hypothesis that There would not exist any relationship of P300 latency in normal hearing- and hearing-impaired adults. Pearson's correlation was found that hearing loss and amplitude of P300 has a strong negative correlation and latency of P300 has a poor positive correlation with hearing loss. Conclusions: Hearing loss affects the P300 measures. Due to hearing loss cognitive decline occurs which results in reduced amplitude and prolonged latency in P300.

The word-with-noise test: test-retest reliability in normal-hearing adults.

To investigate the reliability of the Word-with-Noise Test in a group of normal-hearing adults. Forty-five normal-hearing adult subjects participated in the research. The interval between the first and second assessment was 14 to 28 days, performed during the same time of the day and by the same evaluator. The comparison analysis between the test and the retest was performed considering the general result of the ears, totaling 90 ears evaluated. The inferential analysis included the comparison of the situations in the first and second assessment using the Wilcoxon Test, calculation, and interpretation of the Intraclass Correlation Index. There was a statistically significant difference between the test and retest performances. The intraclass correlation coefficients obtained were indicative of good reliability (r=0.759; p<0.001) for the monosyllabic stimulus and moderate reliability (r=0.631; p<0.001) for the disyllabic stimulus. The Word-with-Noise Test demonstrated satisfactory reliability for both the monosyllabic and disyllabic stimuli.

Teste de palavras no ruído: confiabilidade no teste-reteste em adultos normo-ouvintes

ABSTRACT Purpose To investigate the reliability of the Word-with-Noise Test in a group of normal-hearing adults. Methods Forty-five normal-hearing adult subjects participated in the research. The interval between the first and second assessment was 14 to 28 days, performed during the same time of the day and by the same evaluator. The comparison analysis between the test and the retest was performed considering the general result of the ears, totaling 90 ears evaluated. The inferential analysis included the comparison of the situations in the first and second assessment using the Wilcoxon Test, calculation, and interpretation of the Intraclass Correlation Index. Results There was a statistically significant difference between the test and retest performances. The intraclass correlation coefficients obtained were indicative of good reliability (r=0.759; p&lt;0.001) for the monosyllabic stimulus and moderate reliability (r=0.631; p&lt;0.001) for the disyllabic stimulus. Conclusion The Word-with-Noise Test demonstrated satisfactory reliability for both the monosyllabic and disyllabic stimuli.

Toward Sound Localization Testing in Virtual Reality to Aid in the Screening of Auditory Processing Disorders.

Sound localization testing is key for comprehensive hearing evaluations, particularly in cases of suspected auditory processing disorders. However, sound localization is not commonly assessed in clinical practice, likely due to the complexity and size of conventional measurement systems, which require semicircular loudspeaker arrays in large and acoustically treated rooms. To address this issue, we investigated the feasibility of testing sound localization in virtual reality (VR). Previous research has shown that virtualization can lead to an increase in localization blur. To measure these effects, we conducted a study with a group of normal-hearing adults, comparing sound localization performance in different augmented reality and VR scenarios. We started with a conventional loudspeaker-based measurement setup and gradually moved to a virtual audiovisual environment, testing sound localization in each scenario using a within-participant design. The loudspeaker-based experiment yielded results comparable to those reported in the literature, and the results of the virtual localization test provided new insights into localization performance in state-of-the-art VR environments. By comparing localization performance between the loudspeaker-based and virtual conditions, we were able to estimate the increase in localization blur induced by virtualization relative to a conventional test setup. Notably, our study provides the first proxy normative cutoff values for sound localization testing in VR. As an outlook, we discuss the potential of a VR-based sound localization test as a suitable, accessible, and portable alternative to conventional setups and how it could serve as a time- and resource-saving prescreening tool to avoid unnecessarily extensive and complex laboratory testing.

Automated Speech Audiometry: Can It Work Using Open-Source Pre-Trained Kaldi-NL Automatic Speech Recognition?

A practical speech audiometry tool is the digits-in-noise (DIN) test for hearing screening of populations of varying ages and hearing status. The test is usually conducted by a human supervisor (e.g., clinician), who scores the responses spoken by the listener, or online, where software scores the responses entered by the listener. The test has 24-digit triplets presented in an adaptive staircase procedure, resulting in a speech reception threshold (SRT). We propose an alternative automated DIN test setup that can evaluate spoken responses whilst conducted without a human supervisor, using the open-source automatic speech recognition toolkit, Kaldi-NL. Thirty self-reported normal-hearing Dutch adults (19-64 years) completed one DIN + Kaldi-NL test. Their spoken responses were recorded and used for evaluating the transcript of decoded responses by Kaldi-NL. Study 1 evaluated the Kaldi-NL performance through its word error rate (WER), percentage of summed decoding errors regarding only digits found in the transcript compared to the total number of digits present in the spoken responses. Average WER across participants was 5.0% (range 0-48%, SD = 8.8%), with average decoding errors in three triplets per participant. Study 2 analyzed the effect that triplets with decoding errors from Kaldi-NL had on the DIN test output (SRT), using bootstrapping simulations. Previous research indicated 0.70 dB as the typical within-subject SRT variability for normal-hearing adults. Study 2 showed that up to four triplets with decoding errors produce SRT variations within this range, suggesting that our proposed setup could be feasible for clinical applications.