International Symposium on Auditory and Audiological Research

Udit Saxena¹, Hansapani Rodrigo², Srikanta Mishra^3
1Department of Audiology & Speech Language Pathology, GMES Medical College & Hospital, Ahmedabad, India
²School of Mathematical and Statistical Sciences, The University of Texas Rio Grande Valley, Edinburg, USA
³Department of Communication Sciences & Disorders, University of Texas Rio Grande Valley, Edinburg, USA

Humans can hear up to 20 kHz. Compared to adults, children have excellent hearing in the extended high frequencies (EHFs). Emerging reports suggest that EHFs contribute to speech-in-noise recognition in children who have normal hearing in the standard frequencies (0.25 through 8 kHz) and EHFs (10-16 kHz). However, the effect of EHF hearing impairment in children remains unclear. This case-control study aimed to answer how EHF hearing loss in children is related to cochlear function in lower frequencies and speech-in-noise perception. We measured hearing thresholds in the standard frequencies and EHFs, distortion product otoacoustic emissions, and digit triplets in noise in children (n=542; 4-19 years) with clinically normal audiograms. Thirty-eight children had some degree of impairment (>15 dB HL) for at least one EHF. Children with EHF impairment had relatively higher thresholds in the standard frequencies even though they had clinically normal audiograms. Otoacoustic emissions in the 2-5 kHz region predicted EHF hearing status and were lower for EHF-impaired children than children with no EHF impairment. EHF impairment had a small but statistically significant effect on the speech recognition threshold, when age effects were adjusted using a linear mixed-effects model. There was no effect of otitis media history, although a history of pressure equalization tube surgery excluded from participation. These findings suggest that despite a normal audiogram, EHF hearing impairment is common in children and is associated with pre-clinical cochlear deficits and poor speech-in-noise recognition. The results will be discussed in detail.

Apoorva Prathibha K. S. ¹ , K. V. Nisha¹, Ajith U. Kumar ^1
1 Department of Audiology, All India Institute of Speech and Hearing (AIISH), Naimisham Campus, Manasagangothri, Mysuru, 570006, India

The variability in rehabilitative outcomes of children with ANSD has intrigued Audiologists over past two decades. The current study aimed to compare the hearing aid outcomes in children with ANSD and SNHL. Retrospective data collected from medical reports of 24 participants, aged 1-4.9y was taken for the study. The participants were into two groups based on the patho-physiology of the deficit: ANSD (n = 12, mean age:2.43±1.21 SD) and SNHL (n = 12, mean age:2.43±1.21y SD). While the group equivalency in HA functional benefit (aided hearing aid scores within speech banana) and data at 3 evaluations (baseline, follow up-1, & follow up-2) was considered preliminary criteria for inclusion of both groups, the diagnosis of the former was based on the recommendations of Staar et al. (2000) and the latter group comprised on SNHL children who were age-matched to the former group. The hearing aid benefit was quantified as difference between unaided and aided thresholds of the better ear were obtained at three evaluations for 4 frequencies (500, 1000, 2000 and 4000 Hz). The results of two-way repeated measure ANOVA (4 frequencies × 3 evaluations) revealed significant no main effect of frequency [F(3,66)=0.12, p>0.05] and follow up [F(2,44)=0.61, p>0.05], but the significant effect of group [F(1,66)=0.12, p=0.001] and significant interaction effects [group*followup: F(2,44)=8.31, p=0.001, η_p² =0.03; group*frequency: F(3,66)=5.17, p=0.003, η_p² =0.19; group*frequency*follow up:F(6,132)=4.11, p=0.001, η_p² =0.16] were seen. On post-Hoc independent t-tests, the hearing aid benefit in SNHL group was statistically higher (p<0.05) than ANSD group in 500, 1000 & 2000 Hz frequencies at baseline, and at 500 Hz in follow up 1. This disparity in hearing aid benefit (despite group equivalency age, degree and functional outcomes of hearing aid) between the groups can be traced to the neural asynchrony in ANSD, leading to variable outcomes in them.

Signe Hjorth Fogh^1,2, Shno Koiek^1,2, Tobias Neher^1,2
1Institute of Clinical Research, University of Southern Denmark, Odense, Denmark
²Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark

Recently, a Danish sentence material for assessing speech recognition in noise in school-age children (the ‘børneDAT’ material) was developed. To allow this material to be used clinically, age-specific normative data are required. The aim of the current study was to collect such data for participants aged 6-7, 7-8, 8-9, 9-10, 10-11, 11-12, 12-13 and 20-30 years. Another aim was to assess the test-retest reliability of the collected data. Seventy-four children and 12 adults participated. Speech recognition thresholds (SRTs) were measured twice at two different visits in each of four conditions. In the first two conditions, diotic stationary speech-shaped noise was used. The target speech was presented diotically (N0S0) or interaurally out-of-phase (N0S180). In the other two conditions, two running speech maskers were used. The target speech was presented from 0° and the two speech maskers from either 0° (co-located) or ±90° (spatially separated). In general, the SRTs decreased with higher age, with the groups <10-11 years obtaining higher SRTs than the adults. The SRTs were lowest in the N0S180 and spatially separated conditions (means:  -6.7 and −6.6 dB SNR, respectively) and highest in the co-located condition (mean: −1.8 dB SNR). The within-subject standard deviation of the measurements was smallest in the N0S0 condition (1.0 dB SNR) and largest in the spatially separated condition (1.3 dB SNR). Training effects were on the order of 1-1.5 dB. In conclusion, these results demonstrate the applicability of the børneDAT material for assessing speech recognition in noise in children as young as 6 years.

Shno Koiek^1,2, Jens Bo Nielsen³, Harvey Dillon^4,5, Christian Brandt^1,2, Tobias Neher^1,2
1Institute of Clinical Research, University of Southern Denmark, Odense, Denmark
²Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark
³Hearing Systems, Department of Health Technology, Technical University of Denmark, Lyngby, Denmark
⁴Department of Linguistics, Macquarie University, Sydney, Australia
⁵Manchester Centre for Audiology and Deafness, The University of Manchester, Manchester, United Kingdom.

Several studies have reported negative long-term effects of early-childhood otitis media (OM) on the ability to exploit binaural information for segregating competing speech signals. Whether corresponding effects also manifest themselves at lower levels of auditory processing has not been investigated systematically. The aim of the current study was to investigate the long-term effects of OM on speech recognition in stationary noise or competing speech with or without binaural differences between the target and maskers. Another aim was to investigate the effects of the individual otologic history on these results. Children aged 6-13 years either with (N = 42) or without a history of OM (controls, N = 20) participated. For measurements in stationary noise, the target speech was presented from 0° and the masker from 90°. Speech reception thresholds (SRTs) were then measured with the stimuli presented either binaurally or monaurally to the ear opposite to the noise. For measurements with competing speech, binaural SRTs were measured with the target speech presented from 0° and two speech maskers presented from either 0° or ±90°. For each set of measurements, binaural advantage scores were also calculated. To investigate the influence of the individual otologic history, an index of overall OM severity (OMS) was derived based on OM onset age, overall OM duration, and time since the last OM episode from the children’s otologic records. The OM children showed poorer monaural and binaural SRTs in stationary noise as well as poorer binaural SRTs with speech maskers from ±90° compared to the controls. The binaural advantage scores were similar between the two groups. Consequently, the hearing deficits in noise in OM children appear to arise from lower levels of auditory system. After controlling for age, the defined OMS index was not a significant predictor of the hearing in noise abilities in OM children.

Chuan Wen¹, Nicolai Pedersen¹, Jens Hjortkjær^1,2
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark
²Danish Research Centre for Magnetic Resonance, Centre for Functional and Diagnostic Imaging and Research, Copenhagen University Hospital – Amager and Hvidovre, Copenhagen, Denmark

Acoustic separation of constituent speech streams from multi-talker mixtures is a challenging task for man and machine. Recent studies have shown that visual information of the speaker’s face can aid automatic speech separation systems. Audio-visual speech separation systems based on deep neural networks have achieved high speaker-independent performance with only a single microphone channel. However, large-scale datasets are required to obtain satisfactory results, which introduces a heavy computation overhead. To address this problem, we propose an audio-visual deep-learning-based speech separation framework which decouples the audio-visual fusion process from the separation model. Audio-visual features are first optimized independently using Correlational Neural Networks (CorrNet). Visual features extracted from the fusion stage are subsequently applied to the separation network, constituting ‘visual hints’ at the target speech. For two-talker mixtures, our audiovisual separation framework achieves an average performance of 8.09dB scale-invariant source-to-distortion ratio (SI-SDR) improvement. The performance rivals current state-of-the-art separation systems relying on substantially larger networks and more training data.

Niels Overby¹, Torsten Dau¹, Tobias May^1
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Kgs. Lyngby, Denmark

Dynamic range compression aims to restore audibility for hearing-impaired listeners and is one of the most essential building blocks in modern hearing aids. However, the choice of suitable compression parameters, such as the time constants associated with the level estimation stage, depends on the acoustic conditions, and the perceptual benefit of different parameter configurations is still controversial. Listening tests can provide an accurate assessment of the perceptual effects of compression in a limited set of acoustic conditions, but they are time-consuming and can therefore not be used to optimize the various compression parameters across experimental conditions. While several studies have attempted to link the perceptual outcomes of dynamic range compression to a set of objective metrics, there is no agreement on how to objectively quantify the effects of compression. In the current study, a data-driven distance metric was developed based on objective metrics to analyze different compression systems. This analysis included slow-acting, fast-acting, and ‘scene-aware’ compression that adaptively switched between fast- and slow-acting compression depending on the target source activity. In addition, a reference system termed ‘source-independent compression’ was considered that had access to the individual speech and noise signals. A comprehensive list of objective metrics was considered to evaluate the effect of the different compression systems in a wide variety of acoustic conditions, including both interfering noise and room reverberation. Sparse principal component analysis (PCA) was then applied to derive a compact set of interpretable features that explained the effects of compression as linear combinations of sparsely selected objective metrics. The Euclidean distance, within the reduced dimensionality representation, was used to compare the similarity between the compression systems. This newly developed distance metric allows a systematic analysis and optimization of the parameters of dynamic range compression systems by minimizing the Euclidean distance with respect to the source-independent compression system.

Hendrik Husstedt¹, Wiebke Hilgerdenaar^1,2, Marlitt Frenz¹, Florian Denk ¹, Jürgen Tchorz³, Simone Wollermann^1
1German Institute of Hearing Aids, Lübeck, Germany
²Universität zu Lübeck, Lübeck, Germany
³Technische Hochschule Lübeck, Lübeck, Germany

Hearing aid users often report that short, impulse-like noise signals with high sound pressure levels such as slamming of a door or rattle of dishes are particularly annoying. Although the risk of presenting excessive sound levels is covered by the output limiter of hearing aids, features such as a general noise reduction or the automatic gain control are supposed to have no effect on those impulse signals. Therefore, some hearing aids provide an impulse (or transient) noise reduction that should reduce loud and short noise signals without impairing the desired signals, e.g., speech. In this work, we performed a technical evaluation of the impulse noise reduction of commercially available hearing aids of six different manufacturers. For this purpose, we made anechoic recordings of a set of impulse noises, presented them to the hearing aids attached to a head and torso simulator (KEMAR), and determined the C-weighted peak sound pressure levels at the output of the hearing aids. The hearing aids were fitted to a hearing loss of type N3 (according to IEC 60118-15), and the binaural coupling was activated. Various hearing aid configurations were considered where the general noise reduction, the automatic gain control, the output limiter, and the impulse noise reduction were activated or deactivated. Moreover, the impulse noises were presented in quiet, and within a speech pause of the international speech test signal (ISTS). These results give deep insights into the performance of the impulse noise reduction in hearing aids of multiple manufacturers, for different impulse noises, in quiet and within speech, and the interaction with other hearing aid features.

Ellen Lundorff^1,2,3, Rasmus Skipper⁴, Tobias Neher^1,2, Georg Stiefenhofer^3
1Institute of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark
²Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark
³Scientific Audiology, WS Audiology A/S, Lynge, Denmark
⁴Dept. of Audiology, Odense University Hospital, Odense, Denmark

Sound quality is a critical component for hearing-aid satisfaction and uptake (Kochkin, 2010). A key factor influencing sound quality is the inherent processing delay of hearing aids, which causes audible effects when the processed sound mixes with the direct (un-delayed) sound that enters the ear canal through a dome or vent of an earmold. Depending on the size of the delay, the resulting effect relates to changes in sound timbre, perception of echoes, and altered spatial qualities. Modern hearing aids have delays of approximately 5-10 ms, with 10 ms being the current ‘gold standard’ for what is deemed tolerable (Stone et al., 2008). It is still unclear, however, how short the delay needs to be for ensuring optimal sound quality for the user. This question is of interest because the latest advances in signal processing allow for much shorter delays to be realized in hearing devices. The current study therefore explored the effects of six different hearing-aid delays ranging from 0.2 to 10 ms on perceived sound quality with the help of a realistic hearing-aid simulator. Three types of stimuli and two gain settings were tested. The participants were 10 normal-hearing subjects and 20 subjects with mild-to-moderate sensorineural hearing losses. From the results, it is expected that hearing-aid delay will be inversely related to perceived sound quality, with delays <1 ms resulting in best outcome overall. Hearing loss severity, stimulus type and gain setting are expected to affect this relationship.

Florian Denk¹, Florian Miethling¹, Jürgen Tchorz², Hendrik Husstedt^1
1Deutsches Hörgeräte Institut GmbH, Lübeck, Germany
²Technische Hochschule Lübeck, Germany

While hearing aids provide large benefits for hearing impaired users, it has been repeatedly shown that wearing hearing aids has also detrimental effects on sound localization, spatial hearing and speech perception. Such ‘side effects’ reduce the benefit of hearing aids, and may outweigh them in potential users with a mild or moderate hearing loss, or normal-hearing users of related devices like hearables. This also holds if the hearing aids are adjusted to a transparent setting, i.e., conserve the transmission behavior of the open ear as close as possible given technical limitations like the microphone location, processing delays or self-noise. To gain further insight into the mechanisms impairing speech intelligibility with hearing aids, we equipped normal-hearing subjects with behind-the-ear hearing aids in a transparent setting, and measured the speech intelligibility in different spatial listening situations with noise and speech maskers. These situations included collocated speech and masker, and speech spatially separated from four distributed masker sources, where the direction of the speech source was either static or randomly changed between sentences. The results show generally increased speech reception thresholds in the aided condition as compared to the open ear, a reduced spatial release from masking, and a decreased advantage of the speech masker with respect to the noise masker, particularly in the condition where the speech source direction was randomly varied. Based on an analysis of the KEMAR-recorded stimuli, we discuss which of these results can be attributed to a change in energetic masking effects and disturbed binaural cues, or whether wearing the hearing aids further impaired the segregation of speech and masker on a higher cognitive level.

Michal Feręczkowski^1,2, Benedikte Degn Mikkelsen¹, Tobias Neher^1,2
1Department of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark
²Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark

The dependence of speech intelligibility on the presentation level has traditionally been modelled with non-decreasing functions such as sigmoids, for both normal-hearing listeners and listeners with sensorineural hearing loss (SNHL). A performance decrease at high presentation levels – so-called rollover (RO) in the performance-intensity function – has traditionally been interpreted as a sign of retro-cochlear hearing loss. In some recent research studies based on speech stimuli presented at high levels, RO was also observed in young listeners with normal audiograms (NAs), possibly reflecting synaptopathy. Overall, RO measurements could therefore be a useful tool for characterizing suprathreshold hearing abilities in various listener groups. While RO has been observed in a number of studies that employed monosyllabic words or low-context sentences, reports of RO in the recognition of high-context sentences are lacking. Here, we hypothesized that RO presence in the performance-intensity function is related to the amount of context information available in the employed speech material. To test this hypothesis, 22 young NA adults without any self-reported hearing problems were tested at two presentation levels (80 and 95 dB SPL, broadband). Three speech materials were used: monosyllabic words, low-context sentences and high-context sentences. To avoid ceiling effects and upward spread of masking, all measurements were performed in stationary speech-shaped noise with the stimuli highpass-filtered at 1.4 kHz. Significant RO effects were found in the speech scores collected with the two low-context materials but not the high-context material. Overall, this suggests that low-context sentences allow for sensitive measurements of rollover effects in individual listeners.

Srikar Vijayasarathy¹, Animesh Barman^2
1Department of Audiology, JSS Institute of Speech and Hearing, Mysuru, India
²Department of Audiology, All India Institute of Speech and Hearing, Mysuru, India

The study investigated illusory auditory continuity percept of the vowel /a/, and its relationship with the phonemic restoration of speech in noise in 25 individuals with normal hearing. The perception of continuity for the interrupted vowel /a/ was measured with eight different signal-to-noise ratios of speech-shaped noise (-10 t0 +4 dB SNR in 2 dB steps). Subgroups were identified using cluster analysis (Hierarchical and K-Means). Phonemic restoration in noise was measured for sentences interrupted with silence and -10 dB SNR speech-shaped noise. The correlation between continuity perception and phonemic restoration was investigated. Listeners could be classified into high and low continuity groups based on their continuity rating at poor signal-to-noise ratios. They could be further divided into two subtypes within each group based on the pattern of continuity perception change across signal-to-noise ratios. Correlational analysis indicated that those with a high continuity percept tended to benefit more from phonemic restoration. Individual differences in the dependence on top-down and bottom-up strategies can explain these findings.

Acknowledgements: We are grateful to the participants of the study for their patient co-operation.

Alexander Lunichkin¹, Alisa Gvozdeva¹, Larisa Zaitseva¹, Elena Ogorodnikova², Irina Andreeva^1
1Laboratory of comparative sensory physiology, Sechenov Institute of Evolutionary Physiology and Biochemistry of Russian Academy of Sciences, Saint Petersburg, Russia
²Pavlov Institute of Physiology of Russian Academy of Sciences, Saint Petersburg, Russia

Lombard reflex improves quality of speech communication in noisy environments. The reflex modifies amplitude-frequency parameters of speech, thus, increasing its intelligibility. The aim was to evaluate the spectral changes characterizing the pronunciation of vowels in speech-like noise. Six female Russian native speakers without speech and hearing impairments aged 20-58 years took part in the study. Speech recordings were carried out in an anechoic chamber during single session. Recordings of nine words with [a], [i] and [u] vowels in different stressed positions were made in quiet and in babble-noise of 60 dB(A).  Before the start of the recording, the speaker adjusted the auditory feedback to correspond to its natural level. The speaker pronounced words addressing them to the experimenter to simulate a communicative situation. Spectral characteristics (F0, F1, and F2) were analyzed on stationary intervals of stressed vowels in tool Praat. Wilcoxon Matched Pairs Test was used to compare the individual spectral parameters of vowels in quiet and in babble-noise. It was found that in noise F0 significantly increased compared to quiet for all stressed vowels in all positions. The group-averaged changes amounted to 12%. We also observed a shift in the position of vowels in the F1-F2 two-dimensional vowel space. F1 increased in babble-noise compared to quiet in all cases: for [u] it was 12% (p<0.000001), for [a] – 8% (p<0.0001), for [i] – 14% (p<0.000001). F2 changed in different ways: for [a] it increased by 2% (p<0.01), for [i] it decreased by 1% (p<0.0001). For the vowel [u] F2 was stationary. Note that in noise, the scatter of individual parameters F1 and F2 was less pronounced than in quiet. The data indicates an improvement of articulatory planning and an increase in differences between lombard vowels.

Acknowledgements: The work was supported by State budget (themes no. AAAA-A18-118050790159-4, AAAA-A18-118013090245-6).

Hyojin Kim¹, Viktorija Ratkute¹, Bastian Epp^1
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Lyngby, Denmark

Comodulated masking noise and binaural cues can facilitate detecting a target sound from noise. These cues induce a decrease in detection thresholds, quantified as comodulation masking release (CMR) and binaural masking level difference (BMLD), respectively. Previous studies showed that CMR is affected by a preceding masker, possibly due to the adaptation of the auditory system to features of the preceding masker. However, its relevance to speech perception is unclear. Here, we investigated the effect of the duration of preceding maskers on CMR and BMLD, and its ecological validity using sounds with speech-like spectro-temporal dynamics.
We hypothesized that the adaptation results from top-down processing, and both CMR and BMLD will be affected with increased preceding maskers’ length. We measured CMR and BMLD when the length of preceding maskers varied from 0 (no preceding masker) to 500 ms. We used four different maskers: the reference condition with uncorrelated masker (RR), and three maskers consist of a comodulated masker preceded by three different maskers: uncorrelated masker (RC), comodulated masker (CC), and the masker with comodulated flanking-band (FC). For BMLD, we used interaural phase difference (IPD) of pi. Results showed that CMR was more affected with longer preceding masker only for FC condition while the preceding masker did not affect BMLD. This indicates that grouping of frequencies in preceding masker has influence on following frequency grouping by comodulation.
We further evaluated the ecological validity of such grouping effect with stimuli reflecting formant changes in speech. We set three masker bands at formant frequencies F1, F2, and F3 based on CV combination: /GU/, /FU/, and /PU/. We found that the CMR was little (~2 dB) while BMLD was comparable to previous findings (~9 dB). In conclude, we suggest that there are factors that play a role in frequency grouping in speech other than comodulation.

Paolo A. Mesiano¹, Johannes Zaar^1,2, Lars Bramsløw², Helia Relaño-Iborra^1,3, Torsten Dau^1
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Kongens Lyngby, Denmark
²Eriksholm Research Centre, Snekkersten, Denmark
³Cognitive Systems Section, Department of Applied Mathematics and Computer Science, Technical University of Denmark, Kongens Lyngby, Denmark

The effects of fundamental-frequency (F₀) cues on speech intelligibility in competing-talker scenarios have been investigated extensively. Most related studies have focused on the substantial intelligibility benefit induced by a difference in the average F₀ between a target and an interfering voice when using non-realistic speech materials. However, recent findings suggest that the effect of an average F₀ difference is rather moderate or even negligible when using more realistic speech stimuli. In contrast, it has been suggested that in everyday-life speech, the relevant F₀-related cues might lie in the dynamics of the F₀ trajectories and their differences between competing voices. In this respect, a contrast in F₀ dynamic range between competing talkers has been shown to notably affect the intelligibility of the target speech but has not yet been investigated systematically. In this study, we explored the effect of F₀-dynamic-range contrast between competing talkers in normal-hearing listeners, using Danish everyday-speech type sentences. In order to control the F₀-dynamic-range contrast, we manipulated the acoustic stimuli by expanding or reducing their F₀ dynamic range within the typical variability of natural speech found in the Danish language, thus maintaining the perceived naturalness of the voices. The results of this investigation are discussed and compared to previous studies with the aim to provide a better understanding of the role of F₀ dynamics in competing-talker scenarios.

Malte Wöstmann^1,2, Julia Erb^1,2, Jens Kreitewolf ^1,2,3,4, Jonas Obleser^1,2
1Department of Psychology, University of Lübeck, Lübeck, Germany     
²Center of Brain, Behaviour, and Metabolism, University of Lübeck, Lübeck, Germany;
³Department of Psychology, McGill University, Montreal, Quebec, Canada;
⁴Department of Mathematics and Statistics, McGill University, Montreal, Quebec, Canada

Acoustic noise is pervasive in human environments. Some individuals are more tolerant to noise than others. We here demonstrate the considerable explanatory potential of non-auditory determinants of subjective and objective noise tolerance, namely personality traits neuroticism (being emotionally unstable) and extraversion (being enthusiastic, outgoing). In an online study, we collected demographic information and BIG-5 personality traits in a large, age-varying sample (N = 1,103; 18–74 years). Subjective noise tolerance was assessed with established self-report scales (i.e., Weinstein Noise Sensitivity Scale, WNSS; Speech, Spatial and Qualities of Hearing Scale, SSQ) and a self-adjustment of the maximal tolerable noise level (i.e., Acceptable Noise Level, ANL). Objective noise tolerance was quantified as reception threshold for digit triplets in noise (DTT). In agreement with pre-registered hypotheses (osf.io/fgyj9), higher neuroticism and lower extraversion independently explained lower scores on all subjective noise tolerance tests, while controlling for demographic factors (e.g., age, gender, education, self-reported hearing loss). Interestingly, this pattern reversed for objective noise tolerance, such that higher neuroticism explained lower (i.e., better) speech-in-noise reception thresholds. We quantified the degree to which listeners of different age and personality profiles over- or underrated their own objective noise tolerance. Older age was associated with overrated noise tolerance, characterized by decreasing objective but largely unchanged subjective noise tolerance. Orthogonal to effects of older age, individuals scoring higher on neuroticism underrated their own noise tolerance (i.e., lower subjective than objective noise tolerance), while high-extraversion overrated it. In sum, these results help build a framework for understanding individual differences in noise tolerance and will help tailor future audiological treatment: Personality holds explanatory power for inter-individual differences in coping with acoustic noise, which is a ubiquitous source of distraction and health hazard in human environments.

Acknowledgements: The present work was supported by the International Hearing Foundation (grant to MW and JO) and an ERC Consolidator grant (ERC-CoG-646696 AUDADPT) to JO.

Alex Mepham¹, Yifei Bi², Sven Mattys^1
1Department of Psychology, University of York, York, UK
²College of Foreign Languages, University of Shanghai for Science and Technology, Shanghai, China

Speech-in-speech masking research has identified what is known as release from linguistic masking, i.e., better speech transcription performance against maskers in unknown than known languages. To test whether native (British) and non-native (Mandarin) speakers of English can learn to control the interference of a known language masker, we tracked their ability to transcribe English/Mandarin sentences against English or Mandarin competing talkers over the course of 50 trials. Both native and non-native listeners improved over time. Native listeners exhibited release from linguistic masking, with less masking in Mandarin than English masker conditions. The size of this effect increased over time. In contrast, non-native listeners showed no difference between the two language maskers, and this pattern was constant over time. Masker-to-target intrusion errors decreased over time for native listeners, whereas they were virtually absent for the non-native listeners. These results show that (1) Linguistic masking is worst when the masker language is known to the listener, whether that language is native or non-native, (2) Listeners show worse performance when the masker is the same language as the target speech, and (3) Native listeners become better at suppressing masker interference over time than non-native listeners, which we hypothesize results from reduced spare cognitive capacity in non-native listeners.

Max Væhrens¹, Rodrigo Ordoñez^2
1Acoustics and Audio Technology, Aalborg University, Aalborg, Denmark
²Department of Electronic Systems, Aalborg University, Aalborg, Denmark

Loss of frequency selectivity can affect a person’s ability to understand speech in noisy environments. Therefore, being able to assess frequency resolution can be an important tool in the treatment of hearing disorders. In the present study two indirect measures of frequency selectivity are compared, sensitivity to spectrotemporal modulation (STM) and optimal ratio for distortion product oto-acoustic emissions (DPOAE).  Changes in STM modulation threshold between normal and impaired listeners have been correlated with speech intelligibility in noise and may represent deficiencies related to frequency selectivity and temporal fine structures in impaired hearing. Varying the ratio between the primary frequencies in a DPOAE measurement, an optimal primary ratio for the specific measurement parameters representing the highest DPOAE level can be obtained. The optimal DPOAE ratio decreases with increasing frequency similarly to the equivalent rectangular bandwidth for auditory filters. Changes in the optimal ratio or the lack of a clear maximum may indicate loss of frequency tuning in the basilar membrane (BM). STM sensitivity threshold and optimal DPOAE ratios were measured at test frequencies of 1 and 4 kHz on eleven volunteers with hearing considered normal. The results show no systematic relationships between the two measures for individual subjects. Based on the DPOAE results two groups were identified. Five subjects show well-defined optimal ratios and high DPOAE levels at both tested frequencies, and six subjects show low DPOAE levels at 4 kHz and no clear optimal ratio at 1 kHz. Comparison of the average STM sensitivity thresholds for these two groups only shows a statistical difference with the 4 kHz stimulus with a spectral modulation of 2 cycles per octave.

Raul Sanchez-Lopez^1,2, Gerard Encina-Llamas^1
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Kgs. Lyngby, Denmark
²Interacoustics Research Unit, Kgs. Lyngby, Denmark

A data-driven auditory profiling approach has been recently proposed to provide a more accurate and sensitive clinical evaluation of hearing. Such approach identified four different groups of hearing-impaired listeners that reflected different degrees of two independent types of auditory deficits: speech-intelligibility deficits and loudness perception deficits. Despite the four groups showing differences in several supra-threshold tasks, audiometric thresholds were also significantly different across the four auditory profiles. In the present study, loudness-growth functions obtained in 75 listeners with different hearing abilities and classified as belonging to one of the four auditory profiles (A-D) are presented and discussed. Furthermore, a state-of-the-art computational model of the auditory nerve was used to explore the potential peripheral pathologies that could lead to the loudness-growth functions observed in each of the four profiles. The aims of the study were 1) to evaluate the association between neuronal rate and loudness perception in the four groups of hearing-impaired listeners, and 2) to investigate the role of outer-hair-cell loss and inner-hair-cell loss on the shape of the loudness-growth functions. The experimental results showed that loudness-related deficits were associated with steeper loudness functions and increased hearing thresholds. In contrast, the loudness functions of listeners belonging to profiles with speech-related deficits were shifted towards higher levels. The model simulations were qualitatively similar to the loudness-growth functions obtained in each of the profiles, although the model was only fitted to the audiometric threshold of each profile without any auditory nerve dysfunction. This suggests that abnormal loudness-growth functions can be mainly explained by peripheral processes and that the role of high-frequency inner-hair-cell loss on the excitation patterns may dominate the abnormal loudness perception, especially at high presentation levels.

Acknowledgements: This work was partially supported by the Better hEAring Rehabilitation project (BEAR) Innovation Fund Denmark Grand Solutions 5164-00011B, and UHEAL: Uncovering Hidden Hearing Loss funded by the Novo Nordisk foundation.

Hartmut Meister¹, Moritz Wächtler¹, Josef Kessler², Martin Walger^1,3
1Jean-Uhrmacher-Institute, University of Cologne, Cologne, Germany
²Department of Neurology, University Hospital Cologne, Cologne, Germany
³Clinic of Otorhinolaryngology, Head and Neck Surgery, University of Cologne, Germany

Listening situations with competing talkers pose high demands on both the auditory system as well as cognitive abilities. These “cocktail party situations” can be “static” or “dynamic”, the latter involving the target talker changing in a possibly unpredictable manner (Brungart & Simpson, 2007; Lin & Carlile, 2015; Meister et al., 2020). Here, different attentional mechanisms may play a role. In static cocktail party listening it is assumed that focusing attention on a known target is important. In dynamic listening situations, however, it is also important to monitor several potential targets and to switch attention if the talker of interest changes (Lin & Carlile, 2019; Meister et al., 2020).
In this study we shed light on the underlying mechanisms by analyzing different error types in a listening situation with three competing talkers. Specifically, random errors (omitting or misunderstanding words) and confusion errors (mixing up target and masker) were determined that might give valuable information about the load in static and dynamic cocktail party situations. Moreover, different listener groups (young and older listeners with and without hearing loss) were considered in order to examine potential effects of age and hearing impairment. The presentation discusses how the different errors could reflect effects of auditory stream segregation and factors such as misdirecting attention or losing attentional focus and whether this depends on the listener groups.

Acknowledgements: Supported by Deutsche Forschungsgemeinschaft ME 2751/3-1.

Lindsey N. Van Yper^1,2, Juan Pablo Faúndez², Jaime A. Undurraga^2,3, David McAlpine^2
1Institute for Clinical Research, University of Southern Denmark, Odense, Denmark
²Department of Linguistics, The Australian Hearing Hub, Macquarie University, Sydney, Australia
³Interacoustics Research Unit, Technical University of Denmark, Lyngby, Denmark

Binaural hearing – particularly the ability to process interaural time differences (ITDs) – underpins sound localization and speech perception in noise. Although ITD processing is known to be affected in various clinical populations, it is not standardly assessed in the clinic, primarily because behavioural measures of ITD are time-consuming and difficult to perform. Recent studies have therefore proposed the ‘acoustic change complex (ACC)’ and ‘interaural-phase modulation following responses (IPM-FR)’ as promising techniques for objective evaluation of ITD processing. Here, we determine the optimal stimulus, recording, and analysis parameters for clinical use of these measurements. Results show that reliable ACC and IPM-FR can be obtained from clinically suitable electrode locations (e.g. mastoids referenced to Cz). However, when using the more practical Fpz location as a reference, one may consider removing eye blinks. We also show that stimulus parameters affect IPM-FR and ACC differently, which may suggest that different neural mechanisms are involved in the generation of these responses.

Acknowledgements: This research has been funded by the Australian Government through the Australian Research Council (project number FL160100108).

Pushkar Deshpande^1,2, Christian Brandt^1,2, Stefan Debener³, Tobias Neher^1,2
1Institute of Clinical Research, University of Southern Denmark, Odense, Denmark
²Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark
³Department of Psychology, University of Oldenburg, Oldenburg, Germany

Effective communication requires good speech perception abilities. Speech perception can be assessed with behavioral and electrophysiological methods. Relating these two types of measurements to each other can provide insights into the underlying cortical processes. The current study aimed to develop a digit-based test battery suited for eliciting different speech-evoked cortical responses and to relate the speech-evoked cortical responses to behavioral measures of speech detection, discrimination, and comprehension. Thirty young normal-hearing native Danish speakers with normal or corrected-to-normal vision participated. The digit-triplet lists from the Dantale-I speech corpus were used as stimulus material. All measurements were carried out in the presence of stationary speech-shaped noise at 67 dB(C) SPL. The behavioral measurements included speech detection thresholds (SDTs), speech recognition thresholds (SRTs), and speech comprehension scores (SCSs). For the electrophysiological measurements, multi-channel electroencephalography (EEG) recordings were performed. N100 and P300 responses were evoked using an active auditory oddball paradigm. N400 and Late Positive Complex (LPC) responses were evoked using congruent and incongruent digit sequences that were presented using audio-only or audio-visual paradigms. All EEG components could be evoked successfully. While no correlations between the SDTs and N100 responses were found, the SRTs were correlated with P300 responses (r = -0.45, p < 0.05) and the SCSs were correlated with the EEG responses to the congruent and incongruent digit sequences. Regarding the N400 and LPC responses, there were significant amplitude differences between the audio-only and audio-visual paradigms. Overall, the developed test battery was found to be usable and to produce reliable data. Follow-up studies with hearing-impaired individuals will provide further insights into the consequences of hearing loss on cortical speech processing. The audio-only and audio-visual paradigms will allow investigating similarities and differences between unimodal and bimodal speech perception.

Eric C. Hoover¹, Katherine N. Palandrani^1
1Department of Hearing and Speech Sciences, University of Maryland, College Park, MD, USA

In pure-tone audiometry, threshold is defined as the lowest stimulus level to meet the stopping criteria and not, as in most other psychophysical procedures, as the stimulus level corresponding to a specific probability of detection. A mathematical relationship between audiometric thresholds and the probability of detection remains elusive. This limits our ability to control the audibility of signals amplified by hearing aids and prevents equivalent thresholds from being obtained using other psychophysical methods. Our hypothesis was that the relationship could be established through an analysis of the probability of meeting the stopping criteria at a given stimulus level. A combinatorial solution was obtained for standard audiometric stopping criteria. Results showed that the probability of detection at threshold is maximized at different stimulus levels depending on the events that occur during the test. This suggests that it is possible to relate audiometric thresholds to the probability of detection, but that there are multiple solutions reflecting the multiple possible ways to satisfy the stopping criteria.

Acknowledgements: Work supported by NIH/NIDCD R01 DC015051 to Frederick Gallun.

Palle Rye¹, Dorte Hammershøi^1
1Department of Electronic Systems, Aalborg University, Aalborg, Denmark

We are facing an expected increase in the number of people affected by hearing disorders and an ongoing ambition for improved efficiency, e.g. through expeditious or fewer clinical visits. Recent research suggest that classifying patients according to archetypal profile based on additional supra-threshold diagnostic measures allow a consistent treatment approach targeting the needs and preferences of the specific archetypal profile. The potential benefits of such detailed individual patient profiles are likely to necessitate more diagnostic test time for proper classification. Out-of-clinic diagnostic hearing testing proposes a tempting remedy for the above challenges. However, the lack of in-person guidance during testing may lead to unnecessary confusion or frustration and result in poor reliability proving detrimental to the goal. To counteract such insufficiencies the current study proposes a gamification approach to out-of-clinic hearing diagnostics. The gamification approach focusses on simple, effective instructions and seek to provide intuitive interaction with the diagnostic tests. As an example of this approach, tablet-based implementations of a 3I-3AFC version of Spectro-Temporal Modulation with immediate trial feedback, as well as other supra-threshold measurement methods are demonstrated. Acceptance scores for a sample of the intended patient group are reported using the System Usability Scale. Finally, results are compared to the performance without gamification of

Acknowledgements: This work was supported by Innovation Fund Denmark Grand Solutions 5164-00011B (Better hEAring Rehabilitation project), Oticon, GN Resound, Widex, and other partners (Aalborg University, University of Southern Denmark, the Technical University of Denmark, Force, Aalborg, Odense and Copenhagen University Hospitals). The funding and collaboration of all partners are sincerely acknowledged. a similar patient group.

Michal Feręczkowski^1,2, Tobias Neher^1,2
1Institute of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark
²Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark

In current practice, hearing aids (HAs) are typically fitted based on audiometric thresholds only, even though research suggests that suprathreshold factors play a role for HA outcome. Measurements based on stimuli with linear frequency-specific amplification (‘aided’) carried out at above-conversational levels may provide estimates of suprathreshold hearing deficits that are largely independent of the effects of audiometric thresholds. In some listeners, high presentation levels may lead to a performance decrease, i.e., rollover in the performance-intensity function. Here, we investigated potential links between word recognition scores (WRS) measured at above-conversational levels, uncomfortable levels (UCL) for narrowband noise stimuli, and aided outcome as assessed using the Hearing-in-Noise Test (HINT) as well as the International Outcome Inventory for Hearing Aids (IOI-HA). The participants were 37 experienced HA users with symmetrical, sensorineural hearing losses. Unaided and aided WRS were measured monaurally under headphones with monosyllabic words presented between the most comfortable and uncomfortable level of each participant. The aided HINT measurements were analyzed using a linear mixed-effects model with the unaided and aided WRS, rollover presence, the UCL data, pure-tone average hearing loss, and age as predictors. The IOI-HA scores were analyzed using a multiple linear regression model with the unaided and aided WRS, pure-tone average hearing loss, and age as predictors. The aided WRS data predicted the two outcomes more effectively than any other considered predictor. Additionally, rollover presence was a significant predictor of HINT outcome. Overall, these results imply that suprathreshold deficits as captured by aided WRS measurements performed at higher-than-normal levels can be useful for developing more individualized HA fitting strategies.

Mengchao Zhang¹, Richard Stern², Deborah Moncrieff³, Bharath Chandrasekaran⁴, Catherine Palmer⁴, Christopher Brown^4
1School of Psychology, Cardiff University, Cardiff, United Kingdom
²Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, United States of America
³School of Communication Sciences and Disorders, University of Memphis, Memphis, United States of America
⁴Department of Communication Science and Disorders, University of Pittsburgh, Pittsburgh, United States of America

Recent animal models have suggested that noise exposure can affect supra-threshold temporal envelope (TE) processing without disrupting absolute hearing thresholds due to selective auditory nerve deafferentation, a process which has been named ‘cochlear synaptopathy’. However, evidence of cochlear synaptopathy in human listeners has been inconsistent, sparking debate on the existence of the synaptopathy in humans. This study compares TE processing between young adults enrolled in a dental school and an unexposed control group, because dentistry students are exposed to drill noise that is distinctively high-frequency, and non-traumatic in nature, and the schedule of noise exposure is systematic for students across different years of enrollment. Unvoiced speech recognition in noise modulated at 16 or 32 Hz was chosen to evaluate TE processing. To limit off-frequency contributions and reduce spectral redundancy of the speech, stimuli were band-pass filtered to lower- (0.8 – 2.3 kHz) or higher-frequency regions (1.7 – 12 kHz). The results showed that the group exposed to dental noise performed more poorly than the unexposed for unvoiced speech in modulated noise. The group difference was more robust for the high-frequency speech than low-frequency speech and when the noise was modulated at 32 Hz rather than at 16 Hz. Furthermore, a small but significant difference was found between the two groups when they recognized the speech in spectrally modulated noise for lower-frequency stimuli. Meanwhile, variables such as age, years of musical training, non-dental noise exposure history, and peripheral auditory screening results did not account for a significant amount of variance of the performances. The findings suggest that human listeners with non-traumatic noise exposure could show poor TE processing as was predicted by the hypotheses related to cochlear synaptopathy and that carefully designed measures are needed when exploring cochlear synaptopathy in humans.

Acknowledgements: This study was based on part of the first author’s dissertation works and was supported through funding received from the SHRS PhD Student Award, School of Health and Rehabilitation Sciences, University of Pittsburgh.

Pernille Holtegaard¹, Mercedes C. Duvig¹, Teresa M. C. Gallo¹ , Bastian Epp^1
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, DK-2800, Kgs. Lyngby, Denmark

​Cochlear synaptopathy (CS) has been suggested to predominantly target low spontaneous rate (SR) fibers. The majority of research in CS hypothesize deficits at supra-threshold levels in accordance with this assumption. However, recent work based on auditory models suggests that the loss also includes high-SR fibers, and that these have an off-frequency contribution to coding of moderate-to-high level sounds. This study aims to investigate the integrity of low- and high-SR fibers in CS. Two outcome measures were assumed to reflect AN-fiber integrity, gap detection thresholds (GDTs) and loudness functions. Low presentation levels were assumed to reflect the integrity of on-frequency high-SR fibers, while moderate-to-high levels were assumed to reflect the integrity of low-SR fibers and off-frequency high-SR fibers. The measure of middle ear muscle reflex (MEMR) strength was used as a proxy measure for CS, and a relationship between MEMR and outcome measures was hypothesized. Listeners with high-frequency sensorineural hearing loss (SNHL) with and without noise-induced tinnitus were recruited. It was hypothesized that listeners with tinnitus, a suggested perceptual consequence of CS, would exhibit higher GDTs and shallower loudness functions compared to listeners without tinnitus. GDTs and loudness functions were measured across a range of presentation levels at low frequencies where hearing sensitivity was normal (<25 dB HL). No significant differences were observed between groups for GDTs or MEMR strength. The slope of the upper part of the loudness function at 0.25 kHz was significantly shallower in the group with tinnitus, but no difference was found at 0.5 or 1 kHz. Also, no correlations were found between MEMR strength and the two outcome measures. Overall, the results did not support the hypotheses. However, the results were confounded by small variations in hearing sensitivity, and therefore can neither confirm nor dismiss the role of high-SR fibers in CS.

Acknowledgements: William Demant Foundation.

Sarineh Keshishzadeh¹, Heleen Van Der Biest², Sarah Verhulst^1
1Dept. of Information Technology, Ghent University, 9000 Ghent, Belgium
²Dept. of Rehabilitation Sciences–Audiology, Ghent University, 9000 Ghent, Belgium

Even though age-induced cochlear synaptopathy (CS) has been demonstrated in rodent and human temporal bones, it remains challenging to determine whether this pathology causes speech intelligibility declines. Difficulties in studying the causality of this relationship in humans relate to the necessity to use indirect and non-invasive (often EEG-based) markers of CS. To study the relationship between speech intelligibility and CS in the ageing population, we adopted low- and high-pass filtered speech conditions (Flemish Matrix test in quiet and noise), alongside an extended hearing screening battery that comprised distortion-product otoacoustic emissions, auditory brainstem responses (ABRs), envelope-following responses (EFRs) and extended high-frequency thresholds (EHFTs).

A total of 69 Flemish subjects participated in this study and were divided into two groups: (i) a young control group with normal audiograms (18-25 years) and (ii) Older adults with a normal audiogram, some with complaints of tinnitus or self-reported speech intelligibility problems (45-60 years). We compared our results to a German cohort of 45 listeners that included older participants as well to study how hearing sensitivity, EHFTs and potential biomarkers of CS (EFRs and ABRs) changed as a function of age. The EFR reductions observed in the older group is consistent with age-related CS, and EFR markers were more sensitive to detect individual differences than were ABR markers. At the same time, it remains challenging to determine whether EHF hearing or CS is more important in predicting individual speech intelligibility declines. Based on the model simulations, our EFR marker was only marginally sensitive to outer hair cells deficits, and hence we believe that CS plays an important functional role even though both EHFTs and EFRs were affected by the aging process. We conclude that early markers of sensorineural hearing loss (EFRs or EHFTs) are crucial for a timely diagnosis of speech intelligibility problems with age.

Acknowledgements: This work was supported by the European Research Council (ERC) under the Horizon 2020 Research and Innovation Programme (grant agreement No 678120 RobSpear and No 899858 CochSyn).

Gayathri Kalarikkal¹, G. K. Jayasree¹, Malavika Puthiyadam¹ , Abdul Bahis, K. V. Nisha¹ , Prashanth Prabhu^1
1Department of Audiology, All India Institute of Speech and Hearing, Mysore, India

Individuals with ANSD show multi-faceted ramifications of the disorder which spans many facets of life (emotional, personal and other health related issues) that are not readily apparent in conventional audiological assessment. The aim of the study was to determine the problems and life effects experienced by individuals diagnosed as ANSD. The study followed a cross sectional survey design with 12 adult ANSD participants. Responses to two open-ended questions were classified using the International Classification of Functioning, Disability and Health (ICF) framework. The first question of the two open ended questions for data collection concentrated on the problems related to having ANSD and the second to its life effects. All the responses collected were linked to the corresponding ICF categories according to the ICF’s established linking rules using a simplified content analysis approach (Granberg, 2015). The results of Wilcoxon signed rank test revealed that participants had experienced more difficulties in PQ (/Z/=-2.39, p=0.02, effect size r= 0.36) than LEQ. 70 responses were obtained for problem question, 48 responses for life effects question. Most of the problems and the life effects confronted by individuals with ANSD were linked to activity limitations and participant restrictions (63/119), followed by body function (46/119) and a small number of responses related to environmental factors (9/119). The frequent number of responses related to activity limitations and participation restrictions pertained to “Communicating with—receiving—spoken messages” (d310). For body functions, “hearing functions” (b230), and “general products and technology for personal use in daily living” (e1150) occurred most under environmental factors category. Person with ANSD encounters hearing problem together with life effect problems in terms of activity limitations, participation restrictions, emotional and environmental factors which may hinder their integration to the society. The study emphasizes on person-centred approach providing a holistic model of rehabilitation to the individuals with ANSD.

Kristin Weineck^1,2, Olivia Xin Wen¹, Molly J. Henry^2
1Research Group “Neural and Environmental Rhythms”, Max Planck Institute for Empirical Aesthetics, Frankfurt am Main, Germany
²Institute for Cell Biology and Neuroscience, Goethe University Frankfurt am Main, Frankfurt am Main, Germany

Neural activity in the auditory system can synchronize to the rhythms of sounds, and successful synchronization improves perception. In the context of natural sounds, most previous studies have investigated neural tracking of speech, whereas only a few studies have examined neural responses to natural (polyphonic) music. Similar to speech-tracking studies, music-tracking studies often investigated neural tracking of the music amplitude envelope. We hypothesized that the envelope alone might not best capture the acoustic fluctuations in music that evoke neural synchronization. This study aimed to investigate 1) neural tracking of different music features, 2) tempo-dependence of neural tracking, and 3) the correlation between neural tracking and behavioral responses to music. We conducted an EEG study, where 37 participants listened to music segments (without vocals) at parametrically varied rates (1-4Hz). Each trial consisted of the presentation of one music stimulus (attentive listening, no movement), a partial repetition of the same stimulus (finger tapping to the beat) and behavioral music ratings (enjoyment, familiarity and beat tapping difficulty). We applied converging neural analyses based on 1) temporal response functions and 2) Reliable Components Analysis combined with stimulus–response coherence and correlation. Our results demonstrate that spectral changes in music (“spectral novelty”), as opposed to the amplitude envelope, evoke the most reliable synchronized neural response. Moreover, music with slower beat rates elicited strongest neural synchronization. We also found that neural synchronization was stronger for familiar than for unfamiliar music. Furthermore, a classifier analysis revealed that neural responses to music presented at a single tempo predicted the tempo that the participant would later tap, i.e., whether they tapped the stimulus tempo vs. double or half that rate. Overall, our results indicate that there are tempo-dependent effects on neural synchronization during natural music listening and that spectral fluctuations in music may be critical for communicating the beat.

Acknowledgements: This work was funded by the European Research Council starter grant from Molly Henry.

Hyojin Kim¹, Viktorija Ratkute¹, Bastian Epp^1
1Hearing Systems section, Department of Health Technology, Technical University of Denmark, Lyngby, Denmark

Comodulated masking noise and interaural phase disparity (IPD) can enhance detection of masked signals. Such enhancement in detection performance, or masking release, can be quantified as decreased detection thresholds: comodulation masking release (CMR) and binaural masking level difference (BMLD). While many studies investigated CMR and BMLD, the relevance of masking release at supra-threshold levels is still unclear. Here, we used both psychoacoustic and electro-physiological measures to investigate how CMR and BMLD affect intensity discrimination at supra-threshold levels.

We designed eight masking release conditions where each condition induces different amount of CMR and BMLD. In the psychoacoustic experiment, we investigated whether the difference in the amount of masking release will affect listening at supra-threshold levels. We used intensity just-noticeable difference (JND) to quantify an increase in salience of the tone. For instance, the salience of the tone would increase when the level of the tone is increased by JND. That is, the condition with lower JND would be more salient than the one with higher JND at the same supra-threshold level. As a physiological correlate of JND, we investigated late auditory evoked potentials (LAEPs) with electroencephalography (EEG). We measured P2 at supra-threshold levels from +15 dB to + 25 dB. Our hypothesis was that the increment in P2 with increasing tone level would be inversely proportional to JND.

From the psychoacoustic experiment, we found that JNDs highly depend on the level of the tone. The JNDs were equal at the same intensity of the tone rather than at the same supra-threshold levels. The results from the EEG experiment showed an inverse correlation between the intensity JND and the amplitude of P2. In conclusion, the amplitude of P2 can reflect the salience of masked tone at supra-threshold levels.

Caitlin N. Price¹, Gavin M. Bidelman^2,3,4
1Dept. of Audiology & Speech Pathology, University of Arkansas for Medical Sciences, Little Rock, AR USA
²Institute for Intelligent Systems, University of Memphis, Memphis, TN USA
³School of Communication Sciences & Disorders, University of Memphis, Memphis, TN USA
⁴Dept. of Anatomy & Neurobiology, University of Tennessee Health Sciences Center, Memphis, TN USA

Mild cognitive impairment (MCI) commonly impacts older adults resulting in more rapid cognitive and behavioral declines than typical aging. Individuals with MCI can exhibit impaired receptive speech abilities that may reflect neurophysiological changes in auditory-sensory processing prior to usual cognitive deficits. Benefits from current interventions targeting communication difficulties in MCI are limited. Yet, neuroplasticity associated with musical experience has been implicated in improving neural representations of speech and offsetting age-related declines in perception. Here, we asked whether these experience-dependent effects of musicianship might extend to aberrant aging and offer some degree of cognitive protection against MCI. During a vowel categorization task, we recorded single-channel EEGs in older adults with putative MCI to evaluate speech encoding across subcortical and cortical levels of the auditory system. Critically, listeners varied in their duration of formal musical training experience (0-21 years). Older musicians exhibited sharpened temporal precision in auditory cortical responses suggesting musical experience produces more efficient processing of acoustic features by offsetting age-related neural delays. Additionally, we found robustness of brainstem responses predicted severity of cognitive decline suggesting early speech representations are sensitive to pre-clinical stages of cognitive impairment. Our preliminary results extend prior studies by demonstrating positive benefits of musical experience in older adults with emergent cognitive impairments.

Acknowledgements: This work was supported by grants from the GRAMMY® Foundation and National Institutes of Health (NIH/NIDCD R01DC016267 and R01DC016267-03S1) awarded to G.M.B. Requests for materials should be addressed to G.M.B [gmbdlman@memphis.edu].

Anaïs Bouchet^1,2,  Abigail A. Kressner^1,2, Erik F. Kjærbøl^2, Maaike Van Eeckhoutte^1,2
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Lyngby, Denmark
²Copenhagen Hearing and Balance Center, Ear, Nose, and Throat (ENT) & Audiology Clinic, Rigshospitalet, Copenhagen University Hospital

Functional near-infrared spectroscopy (fNIRS) is a relatively recent imaging technique and presents many advantages for clinical use in audiology: it is silent, compatible with cochlear implants, and child-friendly, and can measure cortical activation in response to acoustic stimuli. Most studies report results at the group level, with the responses averaged across participants. However, for clinical applications, a good test-retest reliability at the individual level is required. Furthermore, a challenge when using fNIRS is to remove physiological confounds from the data, such as the heartbeat and motion artefacts. While various processing methods have been developed, the procedure to follow is not yet standardized. One promising new approach for improving the test-retest reliability is to use short-channel separations. The aim of this study was therefore to investigate individual level test-retest reliability of the fNIRS responses evoked by auditory stimulation for normal-hearing participants. The fNIRS responses of fourteen participants were recorded to silence, a speech-shaped modulated noise, a speech passage, and a music sample. In addition, the speech-shaped noise was presented at two different intensities. The stimuli were presented in a block paradigm. The same testing procedure was repeated with an interval between sessions of one week. Preliminary results show a high variability of test-retest reliability of the fNIRS response across individuals. The results will be further discussed at the conference.

Samuel Smith^1,2, Christian J. Sumner¹, Thom Baguley¹, Paula C. Stacey^1
1NTU Psychology, Nottingham Trent University, Nottingham, U.K.
²Hearing Sciences, University of Nottingham, Nottingham, U.K.

The comprehension of speech, whether for normal hearing or aided, is often supplemented by watching a talker’s facial movements. How do auditory and visual cues combine multiple sources of information to provide a speech benefit? Does audiovisual performance depend only on the unimodal-information, or does the integration process itself vary? Does the integration vary depending on different aspects of the stimulus, or do individuals “integrate differently”? We developed a model based on signal detection theory (SDT) which allows us to test these different possibilities quantitatively. Signal detection theory posits that performance is limited by sensory noise in the signal and internally in the nervous system. The benefits of multiple cues depend on whether internal noise occurs in unimodal processing, or occurs in later processing, after multisensory integration (Micheyl and Oxenham, 2012; J Acoust Soc Am. 131:3970). We propose a model whereby the proportions of unimodal (“early”) and post-integration (“late”) noise can be estimated from unimodal and multisensory performance. This allows us to test whether differences in multisensory performance across experimental variables are best explained by variations in the unisensory performance, or reflect a varying integration process. In previously published data (Stacey et al. 2016; Hear Res. 336:17) we found that SDT provided a good account of audiovisual speech perception, overall. However, previous models were restricted to assuming only one source of noise and neither unisensory nor multisensory noise models predicted the data quantitatively. Our new model, which combines unisensory and multisensory internal noise, fits these data precisely. Furthermore, we find that the integration process shifts towards later multisensory internal noise when temporal fine structure of speech is removed with tone-vocoding. Thus, we can quantify auditory-visual speech perception as an optimal integration of information with multiple sources of internal noise, and the integration itself varies depending on the unisensory signals.

Martha M. Shiell¹, Sergi Rotger-Griful¹, Martin Skoglund^1,2, Johannes Zaar^1,3, Gitte Keidser^1
1Eriksholm Research Centre, Oticon A/S, DK-3070 Snekkersten, Denmark
²Department of Electrical Engineering, Linköping University, Linköping, Sweden
³Department of Health Technology, Technical University of Denmark, Kgs. Lyngby, Denmark

Although hearing aids can be effective at restoring sensory input, some users still struggle to benefit from their devices in real-world communication situations. To better understand and address the needs of these users, we need test paradigms that can capture more ecologically-valid outcome measures. In this presentation, we will describe a range of recent efforts from Eriksholm Research Centre where we explored methods to develop such paradigms, and share recommendations for continued work in this direction. Our efforts spanned seven independent experiments with various groups of hearing-impaired and normal-hearing participants. Over these experiments, we advanced traditional speech intelligibility tests with two additions: (1) stimuli that reflected some real-world complexity, and (2) an accuracy measure aimed capturing the listener’s understanding of the speech material (i.e., comprehension). The stimuli were audiovisual recordings of three unscripted talkers, two of whom engaged in a Diapix task and a third that improvised a monologue. This high level of realism produced the expected challenges to experimental control, but also produced high participant engagement – to the extent that engagement may even have distracted somewhat from the experimental tasks. The accuracy measure was calculated from responses to questions on the speech content. Three styles of questions and their iterations were implemented, in which we targeted the listener’s comprehension while avoiding formulations that rewarded word-recognition strategies. We speculate that, as a side effect of this strategy, our question-response systems evoked increased cognitive abilities for reading, semantic abstraction, and working-memory. Overall, our experiences emphasize the necessity of developing stimuli that are catered to the desired task. Furthermore, they highlight the challenges that are associated with measuring comprehension via extensions of traditional speech intelligibility tests. As such, we suggest that new innovations in behavioural testing may be required for more ecologically-valid outcome measurements.

Acknowledgements: This work was financially supported by the Swedish Research Council (Vetenskapsrådet, VR 2017-06092 Mekanismer och behandling vid åldersrelaterad hörselnedsättninggrant).

Cosima A. Ermert¹, Lu Xia², Brian Man Kai Loong², Janina Fels¹, Sébastien Santurette^2,3
1Institute for Hearing Technology and Acoustics, RWTH Aachen University, Aachen, Germany
²Centre for Applied Audiology Research, Oticon A/S, Smørum, Denmark
³Department of Health Technology, Technical University of Denmark, Kgs. Lyngby, Denmark

Focusing on target signals in noisy environments is a well-known challenge for hearing-impaired listeners. Therefore, a main goal of hearing aid (HA) technology is to enhance target signals and attenuate background noise. An established objective measure to quantify this contrast achieved by the HA is the output signal-to-noise ratio (SNR), commonly estimated using the phase-inversion method (Hagerman and Olofsson, 2004). Output SNR measurements are typically performed in artificial lab setups where target and noise signals are played from discrete directions. However, while such setups can be controlled and modified precisely, they do not adequately represent everyday listening situations. This study investigated the potential and limitations of bringing ecological validity to SNR measurements by using ambisonic reproduction of real sound scenarios as stimuli. Measurements with HAs using different directionality and noise-reduction strategies were performed in multiple pre-recorded 3D scenes. While this allowed a comparison of how the different strategies handled real-life background sounds against specific target sounds, it also introduced challenges to the phase-inversion method. Modern HAs rely increasingly on non-linear processing to adapt to different scenarios, while the phase-inversion method is most accurate under linear processing conditions. Thus, some HA features used for, e.g., feedback management, may induce errors to SNR measurements. The present results underlined the importance of defining an acceptance criterion for the error signal in output SNR measurements. They provided insights on how HA settings must be chosen to eliminate distortions while still measuring the devices at their full potential. When this is considered, measuring HA output signals under realistic conditions can have further useful applications for objective or subjective evaluation. As the increasing complexity of signal processing algorithms is challenging established methods for assessing HAs, it is necessary to discuss which considerations must be made to ensure ecologically valid and future-proof evaluation techniques.

Acknowledgements: The authors would like to thank Jens-Christian Britze Kijne and Boris Søndersted for lab setup, Alexandros Rompelakis for ambisonic sound environment rendering, Asger Heidemann Andersen and Jacob Aderhold for guidance on SNR measurements and post analysis.

Fenja Schwark¹, Stephan D. Ewert¹, Marc René Schädler¹, Volker Hohmann^1,2, Giso Grimm^1,2
1Medizinische Physik, Universität Oldenburg, and Cluster of Excellence “Hearing4all”, Oldenburg, Germany
²HörTech gGmbH, Oldenburg, Germany

In real-time virtual acoustic rendering, the head related directional properties of the receiver, i.e., the listener, are often modeled by convolving the signal with measured head related impulse responses (HRIRs). However, the computational cost of HRIR-convolution is rather high, even when implemented in the spectral domain, and interpolation needs to be applied to simulate all source directions, depending on the spatial resolution of the HRIR catalogues used. In order to reduce the computational cost in low-delay real-time virtual acoustic rendering, this study uses a parameterized digital filter model with delay lines to approximate the direction-dependent features of the head. A data-driven optimization method for the filter parameters is introduced that aims at matching the direction-dependent features of modeled and measured HRIRs using a spectral distance metric. Utilizing an objective binaural speech intelligibility model, it was shown that the speech intelligibility estimate for the optimized model approaches the estimate for the measured HRIRs. This suggests that the parameterized HRIR model may be sufficient to enable plausible spatial perception in virtual acoustic scenes. With the parameterized HRIR model a reduction of computational cost in the order of two magnitudes is possible for virtual acoustic scenes with a small number of objects. Further work will include subjective testing of the model, compared against measured HRIRs.

Acknowledgments: Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Project-ID 352015383 – SFB 1330 B1 and C5.

Moussa Kousa¹, Julien Besle^1
1Department of Psychology, American University of Beirut, Beirut, Lebanon

We evaluated two adaptive methods for fast and reliable measurement of frequency difference limens (DLFs) at multiple base frequencies in naive participants. Our ultimate goal will be to correlate these DLF functions with cortical frequency magnification functions measured in the participants’ primary auditory cortex using fMRI. We first measured DLFs at 8 log-spaced base frequencies [0.2-8 kHz] using the maximum likelihood procedure (MLP; [1]) in 10 participants (2I2AFC task; 5-6 blocks of 30 trials per frequency). The MLP provided unreliable thresholds that varied between consecutive blocks of the same frequency, due to the MLP’s susceptibility to lapses of attention early in a block. Naive participants also found the MLP’s sudden changes in difficulty in the very first trials confusing. We then switched to the Updated Maximum Likelihood procedure (UML; [3]), which has been developed to improve on the MLP’s above weaknesses by simultaneously and adaptively estimating three parameters of the psychometric function (midpoint, slope and lapse rate of attention) and by increasing difficulty more gradually. To verify the validity of the UML thresholds, we compared them to thresholds obtained using the method of constant stimuli (CS) in 14 participants (8S2A task [2]; 8 log-spaced base frequencies between 0.2 and 8 kHz; 2 blocks of 75 trials per frequency per session; session 1: UML, sessions 2&3: UML or CS, counterbalanced across participants). Preliminary results in six participants suggest that the UML and CS procedures yield very similar thresholds, that only 75 trials are necessary to obtain reliable thresholds within a session, but that most participants’ thresholds substantially improved between sessions 1 and 2, suggesting that minimal training remains necessary to obtain accurate thresholds in naive participants

Václav Vencovský^1
1Department of Radioelectronics, Czech Technical University in Prague, Czech Republic

Broadband noise with a sharp falling edge in the power spectrum evokes a monaural noise edge pitch. When comparing with the pitch of a matching pure tone, the same pitch sensation is evoked for pure tone shifted slightly into the spectral region of the noise. In the frequency region between 100 Hz and 2.5 kHz, the pure tone frequency is about 2% to 9% below the noise edge frequency for lowpass filtered noise and about 2% to 20% above the noise edge frequency for highpass filtered noise [Hartmann et al. (2019) J. Acoust. Soc. Am. 145:1993-2008]. The departure from the edge frequency grows as the edge approaches low frequencies. The noise edge pitch can be explained by temporal theories calculating autocorrelation or by a place theory employing lateral inhibition. It is shown here that the noise edge pitch can be also derived from the first order intervals between successive peaks in the temporal fine structure of the noise filtered with a gammatone filterbank. The filterbank was composed of 300 filters distributed between 20 Hz and 10 kHz according to the Cam scale. Histogram of the first-order interpeak intervals reveals a hump centered around the noise edge pitch frequency. The hump is more visually apparent for LP noise than for HP noise as the edge frequency approaches low frequencies (below about 200 Hz), which corroborates experimental results in literature. At 200 Hz, ratio between the pitch frequency estimated as the mid of the hump and the edge frequency is about 1.1, which agrees with literature. These results support temporal theories of pitch perception employing first-order interspike intervals [Huang and Rinzel (2016) Front. Comput. Nerosci. 10:57(1-17)].

Acknowledgements: Supported by an internal grant at the Czech Technical University in Prague SGS20/180/OHK3/3T/13.

Palle Rye¹, Dorte Hammershøi^1
1Department of Electronic Systems, Aalborg University, Aalborg, Denmark

The initial crucial step in hearing rehabilitation is a reliable diagnosis of hearing impairment. Pure tone threshold measurement in a quiet clinical setting is the current gold standard for this purpose. The reliability of pure tone threshold measurement depends on a quiet environment, especially when the threshold of the test subjects are near normal hearing. As a relatively novel procedure for characterizing individual hearing characteristics, ISO 16832 defines a reference method for categorical loudness scaling, where most of the presented stimuli are above the threshold by design. Here the test subject rates the subjective loudness of a set of narrowband auditory stimuli adapted during the test to cover the full dynamic range of the individual. The response is a subjective rating using an 11-point scale from not heard to extremely loud. In the outlined reference method, most of the presented stimuli are above the threshold by design, ideally with only one or two presentations characterized as not heard. The research behind the reference method suggests that the threshold can be reliably predicted by fitting an appropriate model on the collected set of responses, even when the presented set of stimuli are mainly above the threshold. The present study investigates the effects of performing categorical loudness scaling in a noisy environment. Simulations of added background noise to an existing dataset from a clinical setting illustrate the difficulties in estimating auditory thresholds and loudness growth. For test subjects with elevated thresholds, the increase in stimulus level may often be sufficiently above the environmental background noise such that little or no influence is expected. At low audiometric frequencies, however, the predicted hearing threshold is likely to deviate in excess of 5dB for a significant number of patients.

Acknowledgements: This work was supported by Innovation Fund Denmark Grand Solutions 5164-00011B (Better hEAring Rehabilitation project), Oticon, GN Resound, Widex, and other partners (Aalborg University, University of Southern Denmark, the Technical University of Denmark, Force, Aalborg, Odense and Copenhagen University Hospitals). The funding and collaboration of all partners are sincerely acknowledged.

Katrien Vermeire¹, Adrian Fuente^2
1Long Island University, New York City, USA
²Institut Universitaire de Geriatrie de Montreal Research Center, Montreal, Canada

Healthy aging is a priority as the proportion of older adults is drastically increasing. In developed regions the population aged 60 or over is expected to nearly double in 50 years. The prevalence of cognitive decline and dementia increase with age. Age-related hearing impairment (ARHI) is a potential contributor to cognitive decline in older adults and has been shown to relate to poor cognitive performance and dementia. Considering the high prevalence of ARHI and its impact on cognitive performance, there is a need to find protective factors against the effect of ARHI on cognitive performance.

Today, more of the world’s population is bilingual or multilingual than monolingual. Bilingualism has positive effects on cognition and bilingual older adults experience less cognitive decline. However, previous studies on the relationship between bilingualism and cognition have not controlled for ARHI and thus it cannot be determined if bilingualism has a protective effect on cognitive performance in older adults with ARHI.

The aim of this study is to determine if bilingualism offers protection against the negative effect of ARHI on cognition. We hypothesize that bilingual older adults will have better cognitive capacities than their monolingual peers with comparable hearing.

In this study, we will compare cognitive capacities between two groups of older adults (>65 years) with ARHI who do not use hearing aids. Individuals with hearing thresholds in the mild to moderate hearing loss range are considered potential participants. These individuals are to participate in working memory tests, the Reading Span Test and the Corsi Block Tapping test, to assess cognitive abilities. Results are collected and analyzed to determine whether bilingual older adults have better cognitive abilities compared to monolingual older adults with comparable hearing.

Mie L. Jørgensen^1,2, Petteri Hyvärinen¹, Sueli Caporali², Torsten Dau^1
1Hearing Systems Section, Department of Health Technology, Technical University of Denmark, Kgs. Lyngby, Denmark
²WS Audiology, Lynge, Denmark

The objective of the study was to investigate the effect of broadband amplification (125 Hz to 10 kHz) as tinnitus treatment for subjects with high-frequency hearing loss and compare these effects with an active placebo condition using band-limited amplification (125 Hz to 3-4 kHz). The study was a double-blinded crossover study. 22 subjects with a high-frequency (≥ 3 kHz) hearing loss and chronical tinnitus were included in the study and 17 completed the full treatment protocol. Two different hearing aid treatments were provided for 3 months each: broadband amplification that provided gain in the frequency range from 125 Hz to 10 kHz and band-limited amplification that only provided gain in the low frequency range (≤ 3-4 kHz). The effect of the two treatments on tinnitus distress was evaluated with the tinnitus handicap inventory (THI) and the tinnitus functional index (TFI) questionnaires. The effect of the treatment on tinnitus loudness was evaluated with a visual analog scale (VAS) for loudness and a psychoacoustic loudness measure. Furthermore, the tinnitus annoyance was evaluated with a VAS for annoyance. A statistically significant difference was found between the two treatment groups (broadband vs. band-limited amplification) for the treatment-related change in THI and TFI with respect to baseline. Furthermore, a statistically significant difference was found between the two treatment conditions for the annoyance measure. Regarding the loudness measure, no statistically significant differences were found between the treatments, although there was a trend towards lower VAS-based loudness measure resulting from the broadband amplification. No changes were observed for the tinnitus pitch between the different conditions. Overall, the results from the present study suggest that, tinnitus patients with high-frequency hearing loss can experience a decrease in the tinnitus related distress, annoyance and loudness from high-frequency amplification.

Lukas Jürgensen^1,2, Hendrik Husstedt², Florian Denk^2
1Universität zu Lübeck, Lübeck, Germany
²German Institute of Hearing Aids, Lübeck, Germany

As medical devices, hearing aids compensate for a hearing loss but also include several interacting features such as directional microphones, noise reduction or feedback reduction. Detailed knowledge and access to these features would be beneficial for researchers. However, the amount of insight and control into the signal processing strategies of commercially available hearing aids is often quite limited. One approach that could possibly be used as a research hearing aid, is the open Master Hearing Aid (openMHA), which has recently been made available as an open-source tool. In combination with the Portable Hearing Lab (PHL), a portable miniature computer with the opportunity to connect realistic hearing aid headsets, the openMHA can be seen as a full hearing aid prototype. In this contribution, the frequency dependent gains, latencies, the performance of the feedback reduction and the characteristics of the hearing aid channels of the PHL equipped with BTE-RIC headsets running a standard openMHA configuration, were measured in a test box setup, and partly in a KEMAR setup as well. The results measured in the test box setup were also compared to those achieved with two technology levels of a recent commercial hearing aid series. The results show functional feedback reduction for both the prototype and the commercial hearing aids. The latencies of the prototype were 4 ms longer, and the maximum provided gain was 10 dB higher compared to the commercial devices. The results regarding the hearing aid channels show different strategies used in the tested devices without indicating a better or worse performance in either of them. All in all, the hearing aid prototype performed on a comparable level as the commercial hearing aids for the assessed metrics and thus can be viewed as a fully functional hearing aid for research purposes.

Brian K. Man¹, Elaine H. N. Ng^1,2, Emina Alickovic^3,4
1Oticon A/S, Smørum, Denmark
²Department of Behavioral Sciences and Learning, Linkoping University, Linkoping, Sweden
³Eriksholm Research Centre, Oticon A/S, Snekkersten, Denmark
⁴Department of Electrical Engineering, Linkoping University, Linkoping, Sweden

Everyday listening situations that require listeners to selectively attend to a talker of interest in noisy environments with multiple competing talkers are among the most challenging situations encountered by hearing-impaired listeners. Such challenges become even more pronounced with increasing background noise level and may partially be overcome by adequate hearing aid (HA) amplification and noise reduction (NR) support. Using Electroencephalography (EEG), it has been demonstrated that the auditory cortex selectively represents the target talker with significantly higher fidelity than other competing talkers in normal-hearing and hearing-impaired listeners. An NR scheme in commercial HAs was also found to enhance the neural representation of the foreground and entire acoustic scene in early (<85 ms) EEG responses and to enhance neural representations of the target and masker speech and suppress the neural representation of the background noise in late (> 85ms) EEG responses. Motivated by these findings, in this review we investigate whether neural representation of speech in distinct hierarchical stages is affected by NR scheme and configuration in commercial HAs. We show that both the choice of NR scheme and configuration can significantly affect the neural speech processing. By using auditory-attention decoding methods, we show that selecting the adequate NR configuration leads to significantly better representation of the entire acoustic scene and foreground in early EEG responses and to significantly better representation of target and masker speech in late EEG responses in noisy environments. These results suggest that EEG-based auditory attention decoding methods may be sensitive to the choice of HA signal processing configurations in HA users.

Kristin Ohlmann¹, Florian Denk,² Birger Kollmeier^1
1Medical Physics, University of Oldenburg, Oldenburg, Germany
²German Institute of Hearing Aids, Lübeck, Germany

Acoustic transparency – i.e., the ability of a hearing device to “just amplify without side effects” by providing an overall transfer function to the eardrum mimicking the open ear – is a desirable feature to increase the spontaneous acceptance. However, an individual equalization filter accounting for individual ear acoustics is difficult to achieve in real hearing devices, because the required individualized acoustic transfer functions are not available for most applications unless measurements at the eardrum are performed. Furthermore, the processing latency and the leakage of external sound into the ear canal can deteriorate the perceived sound quality. We therefore assessed how sound quality is influenced by different parameters, such as using individually measured or generic data from a dummy head to compute the equalization filter, as well as different latencies or vent configurations. Also, we evaluated how a direction-independent approximation of the direction-dependent transfer function from hearing device microphone to eardrum can be obtained, which allows for a good estimation of the signal at the open ear. Based on individually measurement data, a hearing device was simulated in 10 normal hearing participants’ ears using individual binaural synthesis, allowing to freely vary all hearing device parameters. Acoustic scenes “heard” through the different virtual hearing devices were presented in a MUSHRA-like framework via headphones, and the overall sound quality was rated by the participants. Quality ratings for conditions with individualized transfer functions tend to be higher than for generic data, while latency and leakage show no large overall influence. Differences tend to be larger for complex scenes. In addition to the subjective listening tests, objective quality measures were obtained through a set of models. A good agreement between subjective and objective data was achieved using the GPSMq model. This will support the interpretation and parameter optimization for acoustic transparency in the future.

Sabina S. Houmoeller^1,2, Anne Wolff³, Li-Tang Tsai¹, Sreeram K. Narayanan⁵, Dan D. Hougaard^3,4, Michael Gaihede^3,4, Tobias Neher¹, Christian Godballe^1,2, Jesper H. Schmidt^1,2
1Research Unit for ORL – Head & Neck Surgery and Audiology, Odense University Hospital, Odense, Denmark; University of Southern Denmark, Odense, Denmark
²OPEN, Odense Patient data Explorative Network, Odense University Hospital, Odense, Denmark
³Department of Otolaryngology, Head & Neck Surgery and Audiology, Aalborg University Hospital, Aalborg, Denmark
⁴Department of Clinical Medicine, Aalborg University, Aalborg, Denmark
⁵Department of Electronic Systems, Aalborg University, Aalborg, Denmark

Independent research supporting the choice of hearing aid technology level is lacking. Thus, the main purpose of this study was to explore reported outcomes for older adults with presbycusis using premium-feature and basic-feature hearing aids. Secondly, we investigated if differences in gain prescription measured with real-ear measurements explain differences in self-reported outcomes. The study was designed as a randomized controlled trial in which 190 first-time hearing aid users (≥ 60 years) with symmetric bilateral presbycusis were allocated to either a premium-feature or basic-feature hearing aid. The randomization was stratified on age, sex, and word recognition score. The outcomes were designed to assess their perceived hearing abilities and the effectiveness of hearing aids. Two types of self-reported questionnaires were used: the International Outcome Inventory for Hearing Aids (IOI-HA) and the short form of the Speech, Spatial, and Qualities of Hearing Scale (SSQ-12) questionnaire. In addition, insertion gain at first-fit were measured for all fitted hearing aids. Premium-feature hearing aid users reported 0.7 (95%CI: 0.2;1.1) scale points higher overall SSQ-12 score per item compared to basicfeature hearing aid users. Differences in the prescribed gain at 1 and 2 kHz were observed between premium and basic hearing aids within each company but did not explain the differences in reported outcomes. No statistically significant difference in reported hearing aid effectiveness between the two levels of technology was found. Overall, this study found evidence that premium-feature devices yielded better self-reported outcomes than basicfeature devices.

Acknowledgements: This research was funded by Innovation Fund Denmark Grand Solutions 5164-00011B (‘BEAR project’), GN Hearing, Oticon and WS Audiology. The collaboration with other partners (Aalborg University, Force as well as the university hospitals in Odense, Copenhagen and Aalborg) is sincerely acknowledged.

Sreeram K. Narayanan¹, Anne Wolff³, Sabina S. Houmoeller^4,5, Li-Tang Tsai⁵, Dan Hougaard^2,3, Michael Gaihede^2,3, Jesper H. Schmidt^4,5, Dorte Hammershøi^1
1Department of Electronics System, Aalborg University, Aalborg, Denmark
²Department of Otolaryngology, Head and Neck Surgery, Aalborg University Hospital, Aalborg, Denmark
³Department of Clinical Medicine, Aalborg University, Aalborg, Denmark
⁴Department of Oto-rhino-laryngology, Odense University Hospital, Odense, Denmark
⁵Institute of Clinical Research, Faculty of Health Sciences, University of Southern Denmark, Odense, Denmark

Adjustments to hearing aid (HA) are customarily performed according to patients’ preference either during or after fitting. The adjustments can be either functional or peripheral. Some patients seek immediate help due to a higher degree of challenges or dissatifisfaction with HA performance. However, most do not seek help in due time, which could delay HA rehabilitation. The adjustment timeline and the type of adjustments performed can provide insight into the patients’ behavior and preferences. It would also be worthy of understanding the relationship between the time of adjustments and self-reported outcomes. In the Better hEAring Rehabilitation (BEAR) project, information about adjustments performed within the first two months of initial fit, during two months follow-up visit, and any adjustments thereafter and within one year of fitting were collected using a non-standardized questionnaire. Data from 617 HA users who answered all the questions in the abbreviated version of the Speech, Spatial, and Quality of Hearing (SSQ-12) questionnaire and the International Outcome Inventory for Hearing Aids (IOI-HA) gathered during two months follow-up visit and one-year follow-up, were included in the analysis. The users were divided into eight groups according to the time of HA adjustment. Results show that most of the patients got their hearing aid adjusted during the follow-up visit, indicating a high prevalence of intention or the need for HA adjustment when given an opportunity. This is supported by fewer people adjusting the HA after the two-month follow-up visit. Patients who had their HAs adjusted at various timelines reported significantly lower IOI-HA Factor 1, IOI-HA Factor 2, and SSQ speech domain scores than patients who had no adjustments during the first year of use with the initial fit. Thus, the study establishes the relation between HA adjustments and self-reported outcomes, emphasizing the importance of a correct initial fit and follow-up in rehabilitation.

Acknowledgements: Collaboration and support by Innovation Fund Denmark (Grand Solutions 5164-00011B); Oticon, G.N. Hearing, Widex-Sivantos Audiology and other partners (Aalborg University Hospital, Odense University Hospital, Aalborg University, Technical University of Denmark, FORCE Technology; and, Copenhagen University Hospital) is sincerely acknowledged.

Jacques Grange¹, John Culling^1 
1School of Psychology, Cardiff University, Cardiff, United Kingdom

Pathologies underlying sensorineural hearing loss (SNHL) cannot yet all be differentially diagnosed. We are developing means of pathology discrimination with an advanced SNHL simulator. The physiologically inspired model of the auditory periphery (MAP, Meddis et al., 1986~2018) simulates stimulus encoding at the auditory nerve (AN) level through the firing patterns of 30,000 fibers arranged across 30 best frequencies (56 Hz to 8 kHz) and 3 spontaneous rates. Such encoding is then decoded/converted back into an acoustic signal to be presented to young, normally hearing listeners in psychophysical tasks. Simulator validation was obtained with a speech-in-noise intelligibility task for which simulated-normal-hearing speech reception thresholds (SRTs) were just 1 dB higher than those obtained for unprocessed stimuli. By inserting specific pathologies in the model, we believe one can reveal their psychophysical signatures. We first demonstrated the importance of efferent reflexes to the faithful coding of the temporal modulations that carry speech information. With both reflexes disabled, SRTs grew by 3-4 dB. Simulated AN rate-level functions illustrate how efferent reflexes enable the AN dynamic-range adaptation to context level that prevents information loss. While deactivating 70% of AN fibers or halving the endocochlear potential (EP) did not lead to any appreciable SRT inflation, total outer haircell (OHC) knockout led to a smaller SRT inflation than that found when efferent reflexes were disabled. Circa 90% general deafferentiation was required to reflect performance found in hearing impaired listeners, an outcome consistent with stochastic under-sampling predictions by Lopez-Poveda and Barrios (2013). Both ITD discrimination and MAA thresholds exhibited near-linear growth with the log of remaining efferent fibers. With decreasing EP or OHC count, SRTs inflated and dip-listening and F0-segregation benefits decreased. However, such benefits did not decrease with up to 96% deafferentation, despite SRT inflation. Overall, our simulator paves to way to enabling differential diagnosis of SNHL.

Acknowledgements: This work was funded by EPSRC.

Harshavardhan Settibhaktini¹, Michael G. Heinz², Ananthakrishna Chintanpalli^1
1Department of Electrical and Electronics Engineering, Birla Institute of Technology and Science, Pilani Campus, Rajasthan 333031, India.
²Department of Speech, Language and Hearing Sciences, and Weldon School of Biomedical Engineering, Purdue University, West Lafayette, Indiana 47907-2028, USA.

Listeners hear the speech sounds with varying durations in a real-world environment. A fundamental frequency (F0) difference is an important cue that younger adults with normal-hearing utilize to segregate simultaneously presented multiple speeches. Concurrent vowel identification is studied to understand the effect of F0 difference in identifying two simultaneously presented vowels. Behavioral studies on concurrent vowel identification indicate that the ability to utilize an F0 difference cue is reduced when the stimulus-duration is reduced to 50-ms from 200-ms. Using the computational modeling, Settibhaktini and Chintanpalli [2020, Speech Commun.] showed that the lower concurrent-vowel scores across F0 differences for 50-ms could be predicted by limiting the F0-guided cue for segregation in a modified Meddis and Hewitt algorithm [1992, JASA]. Behavioral studies on concurrent vowels, with longer (>200-ms) duration, indicate that older adults had difficulty in availing F0 difference cue for identification, when compared with younger adults. To understand the age effect, Settibhaktini et al [under review, JASA] developed an older normal-hearing model by cascading a physiologically realistic auditory-nerve model [Bruce et al, 2018, Hear. Res.] with a modified Meddis and Hewitt algorithm to predict the concurrent-vowel scores. The peripheral model included endocochlear-potential loss and cochlear synaptopathy. The model scores were successful in capturing the reduced scores across F0 difference as observed in older adults. The goal of the current study is to predict the concurrent-vowel scores for older adults with shorter duration. The same older normal-hearing model was used to predict the concurrent-vowel scores but with a limited F0-guided cue in a modified Meddis and Hewitt algorithm. Our preliminary results suggest that older adults had larger difficulty in availing the F0 difference cue that resulted in reduced concurrent-vowel scores, than that of younger adults. These predictions can motivate future behavioral study in shorter duration, for which the data is currently unavailable.

Acknowledgements: This work was supported by the third author’s Outstanding Potential for Excellence in Research and Academics Grant (FR/SCM/160714/EEE) and Research Initiation Grant (no. 68), awarded by BITS Pilani, Pilani campus, Rajasthan, India and by the second author’s NIH Grant (R01-DC009838).

Mengchao Zhang¹, Jacques Grange¹, John Culling^1
1School of Psychology, Cardiff University, Cardiff, United Kingdom

Cochlear synaptopathy is a selective loss of auditory nerve fibers with low spontaneous rates (SR) after noise exposure or aging, which is thought to contribute to hidden hearing deficits, especially the ability to process suprathreshold temporal envelopes (TEs). However, evidence of cochlear synaptopathy in humans is unclear due to difficulty in documenting noise exposure history and selecting sensitive measures. The present study uses a computational model to simulate and examine the impact of cochlear synaptopathy on TE perception. Auditory nerves from different SR classes were selectively deactivated in a physiologically inspired auditory model, and then the neural signals of the model were decoded into soundwaves for perceptual evaluation. Simulated synaptopathy (deactivating low-SR fibers) was compared to a normal condition, a more severe version of cochlear synaptopathy (deactivating low- and medium-SR fibers), and a loss of high-SR fibers. TE perception was evaluated through amplitude modulation detection, speech recognition in modulated noise, and recognition of unvoiced speech in modulated noise. Overall, cochlear synaptopathy impaired TE perception, but deactivating high-SR fibers showed no significant difference from the normal condition. The severity of impact of synaptopathy differed based on the task parameters. The modulation detection threshold difference between the normal and the synaptopathy conditions decreased from about 13 dB at 16 Hz to about 9 dB at 64 Hz. For speech tasks, loss of low-SR fibers alone degraded the speech recognition threshold compared to normal conditions by about 1 dB for natural speech but about 4.6 dB for unvoiced speech. In summary, the simulation supports the theoretical role of low-SR fibers in coding suprathreshold TE and shows that sensitive TE measures for cochlear synaptopathy requires a careful selection of task. 

Acknowledgements: The study is supported through the EPSRC grant (Grant No.: EP/R010722/1). 

Piotr Majdak¹, Clara Hollomey¹, Robert Baumgartner^1
1Acoustics Research Institute, Austrian Academy of Sciences, Vienna, Austria

The auditory modeling toolbox (AMT) is a Matlab/Octave toolbox for the development and application of auditory computational models with a particular focus on binaural hearing. The AMT aims at reproducing model predictions and at providing user-friendly access, allowing students and researchers to work with and to advance existing models. To this end, it consists of implementations of auditory models, structured in-code documentation, and the auditory data required to run the models. Model implementations can be evaluated by running so-called demonstrations which are quick presentations of a model and by starting so-called experiments aiming at reproducing results from the corresponding publications. With its version 1.0, the AMT provides a sophisticated framework including caching mechanisms, online repositories, general purpose functions, and plotting functionality, all intended to encourage the enhancement of existing models. For future contributions, the AMT offers multi-licensing of the model implementations, clear display of authorship, and citations to their authors’ publications. The AMT 1.0 includes over 50 models and is freely available as an open-source package from http://www.amtoolbox.org.

Acknowledgements: Partially funded from the European Union’s Horizon 2020 research and innovation programme, project SONICOM, under grant agreement No 101017743.

Christopher Bergevin^1 
1Department of Physics & Astronomy, York University, Toronto, Canada

The auditory periphery is comprised of many disparate elements actively working together to allow for high sensitivity and selectivity. For example, the eardrum appears to move a fraction of the atomic diameter of hydrogen in response to incident sounds at threshold. To achieve such sensitivity, especially in light of thermal noise, biophysical cooperation amongst the elements of the ear is crucial. We argue here that a key principle governing those complex interactions is synchrony. By this we mean the dynamics associated with weakly-coupled self-sustained (i.e., active) oscillators. Here we describe a non-mammalian model, the Anolis lizard, to elucidate how synchrony manifests concurrently at different levels of the periphery. First, within a given inner ear, evidence suggests hair cells metabolically use energy to behave as limit cycle oscillators. Further, they couple together to form groups (or “clusters”) that synchronize, effectively allowing them to increase their sensitivity and selectivity to low-level sounds. Second, by virtue of direct coupling between the tympanic membranes via the interaural canal, the two active ears can also synchrnonize, possibly allowing for improvements in the localization to sounds close to threshold. In parallel, lizards offer an opportunity to explore how synchrony change across the lifespan, given the ability of hair cell regeneration to repair damage to the sensory epithelium. An overarching goal is to explore how these results elucidating cooperativity might be generalized more broadly to the mammalian auditory pathway (e.g., feedback loops in cortical networks). 

Yu Zhao¹, Yumeng Zhang¹, Houguang Liu¹, Xinsheng Huang^2 
1School of Mechatronic Engineering, China University of Mining and Technology, Xuzhou, P. R. China.
²Department of Otolaryngology, Zhongshan Hospital affiliated to Fudan University, Shanghai, P.R. China.

In order to study the effect of the actuator stimulating position on the performance of bone conduction hearing aids, the bone conduction sound transmission mechanism was studied. To facilitate this study, a coupled a finite element (FE) model of the human head and the ear was built. Firstly, a series of micro-CT images of a human head was used to establish an FE model of the human head, and the reliability of the model was verified by comparing the model-predicted mechanical impedance of the excitation position and the acceleration response of the promontory with the experimental data. Secondly, an FE model of human ear, which consists of the ear canal, the middle ear, and the spiral cochlea incorporating the cochlear third windows was also established and verified. Then, based on these two FE models, we constructed a coupled FE model of the human head and the ear via coupling the corresponding nodes, and applied the force excitation at different sites on the model’s head. Finally, through comparative analysis of the dynamic response characteristics of the basilar membrane in the coupled FE model, the effect of different excitation positions of the actuator on the hearing compensation performance of the bone conduction hearing aid was studied. The results show that the best excitation position for the bone conduction’s actuator is near the cochlea, especially the one close to the mastoid. Besides, stimulating the human head close to the cochlea would improve the bone conduction’s performance at low-mid frequencies.

Acknowledgements: This work was supported by the National Natural Science Foundation of China (grant number 51775547).

Francis X. Smith¹, Bob McMurray^2,3, Ruth Litovsky⁴, Inyong Choi^1,3
1Department of Otolaryngology, University of Iowa, Iowa City, United States
²Department of Psychology and Brain Sciences, University of Iowa, Iowa City, United States
³Department of Communication Sciences and Disorders, University of Iowa, Iowa City, United States
⁴Department of Communication Sciences and Disorders, University of Wisconsin-Madison, Madison, United States

The normal auditory system has remarkable mechanisms that integrate sounds from the two ears, enabling listeners to utilize binaural information for achieving critical tasks in their everyday communication. Such binaural benefits include localization (that depends on the computation of binaural disparity) and detecting weak sounds (based on binaural redundancy processing and summation). For patients with single sided deafness (SSD), a cochlear implant (CI) helps to improve sound localization and speech recognition in some noisy scenarios. However, it is not yet known whether CI recovers binaural summation mechanisms for detecting weak sounds in SSD listeners. In this study, the binaural benefit of soft-speech recognition was tested in six SSD CI patients and thirty normal hearing (NH) listeners. SSD CI patients were tested in a four-alternative forced choice speech recognition task using soft speech in three conditions: using only their acoustic ear, only their electrical stimulation ear, or using both ears. NH subjects performed the same speech recognition task in four conditions: one ear with normal speech, one ear with noise vocoded speech, a combination of normal speech to one ear and vocoded speech to the other, or normal speech played to both ears. Electroencephalographic recordings were obtained from all participants as they completed the task. As expected, we observed an accuracy benefit for speech recognition when listening with both ears, even when the input to one ear is CI electrical stimulation / noise vocoded compared to listening with only one ear. Surprisingly, however, the early auditory response observed in the N1P2 complex was lower in amplitude in the combined conditions compared to listening with either one or two normal acoustic inputs. This suggests that the binaural summation benefit in SSD CI users is not solely due to enhanced encoding during early auditory processing, but due to late cognitive processes.

Jean Hong¹, Phil Gander², Joel Berger², Timothy Griffiths³, Inyong Choi^1,4
1Department of Otolaryngology – Head and Neck Surgery, University of Iowa Hospitals and Clinics, Iowa City, USA 
²Department of Neurosurgery, University of Iowa Hospitals and Clinics, Iowa City, USA 
³Biosciences Institute, Newcastle University, Newcastle upon Tyne, UK  
⁴Department of Communication Sciences and Disorders, University of Iowa

Most cochlear implant (CI) users struggle with understanding speech in noisy environments. This challenge may be due to difficulty in separating the target stream from the mixture of sounds. Our previous work demonstrated that figure-ground perception, the ability to detect a ‘figure’ consisting of temporally-coherent tones among less-coherent ‘background’ components, is a strong predictor for successful speech-in-noise perception in normal hearing listeners (Holmes & Griffiths, 2019). However, it is unclear if this temporal coherence-based grouping mechanism predicts CI users’ speech-in-noise ability as well. To address this question, we recruited forty-seven CI users who completed speech-in-noise and figure-ground tasks. In the figure-ground task a sound complex consisting of a figure component that was either fixed (present) or random (absent) in frequency was played simultaneously with a background component of random tones; the participant was asked to detect the presence of the figure component. To isolate the contribution of electric hearing to figure detection we band-passed the figure component from 1-8 kHz and ensured half octave separation among the coherent frequencies to reduce cochlear implant channel interaction. The speech-in-noise task consisted of single words presented in multi-talker babble (4AFC). The results from across-subject correlation analysis indicated that successful figure-ground performance was also a strong predictor for successful speech-in-noise performance among CI users. This research can provide new insights on how to better assess speech-in-noise deficits among CI users.

Meisam Arjmandi^1,2, Julie Arenberg^1,2 
1Department of Otolaryngology-Head and Neck Surgery, Harvard Medical School, Boston, MA, USA
²Audiology Division, Massachusetts Eye and Ear, Boston, MA, USA

Psychophysical tuning curves (PTCs) are effective measures to evaluate frequency selectivity in listeners with cochlear implants (CIs). We are developing a fast and reliable method to measure 5-point PTCs using acoustic stimulation via Bluetooth in CI listeners. The procedure reduces the number of measurements required by eliminating the time to measure thresholds and most comfortable listening levels for each electrode. The pure tones used for masker and probe signals matched the corner frequencies of the analysis filter bands in the processor settings such that mainly one electrode is stimulated. Rather than adaptively varying the level of the masker tone, the masker level is fixed at 56 and 46 dB SPL. Using a forward masking paradigm, the masker frequency is adaptively varied across electrode/frequencies using a two-alternative, two-interval forced choice procedure and frequency thresholds are obtained for the low and high frequency sides, relative to the probe electrode. The tip of the PTC is measured by on-frequency masking, adaptively varying the masker level. Stimuli were controlled through a custom Matlab interface with the AFC software in the background (Ewert, 2013). Data with this method was compared with the gold standard, direct electrical stimulation procedure (Kreft et al., 2019). Preliminary data from two adult CI listeners showed morphologically similar PTCs between acoustic stimulation via Bluetooth and direct electrical stimulation. We characterized PTCs using the apical, basal, and mean slopes. The measures of PTCs calculated from the Bluetooth and direct stimulation procedures were highly correlated for the apical and mean slopes (Pearson r>.9, p<.5). This new tool may be useful for rapid and reliable measurement of PTCs to evaluate frequency selectivity for individual electrode sites. Implementation is for Advanced Bionics devices, but it could be modified for other manufacturers. Developing patient-specific strategies based on PTCs to determine optimal fitting is becoming more feasible.

Acknowledgements: This work was supported by the NIH National Institute on Deafness and Other Communication Disorders Grant RO1 DC012142 (JGA).