Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jan 1.
Published in final edited form as: Hear Res. 2022 Nov 13;427:108649. doi: 10.1016/j.heares.2022.108649

Differences in neural encoding of speech in noise between cochlear implant users with and without preserved acoustic hearing

Hwan Shim 1, Subong Kim 2, Jean Hong 3, Youngmin Na 4, Jihwan Woo 5, Marlan Hansen 6, Bruce Gantz 6, Inyong Choi 3,6,*
PMCID: PMC9842477  NIHMSID: NIHMS1854398  PMID: 36462377

Abstract

Cochlear implants (CIs) have evolved to combine residual acoustic hearing with electric hearing. It has been expected that CI users with residual acoustic hearing experience better speech-in-noise perception than CI-only listeners because preserved acoustic cues aid unmasking speech from background noise. This study sought neural substrate of better speech unmasking in CI users with preserved acoustic hearing compared to those with lower degree of acoustic hearing. Cortical evoked responses to speech in multi-talker babble noise were compared between 29 Hybrid (i.e., electric acoustic stimulation or EAS) and 29 electric-only CI users. The amplitude ratio of evoked responses to speech and noise, or internal SNR, was significantly larger in the CI users with EAS. This result indicates that CI users with better residual acoustic hearing exhibit enhanced unmasking of speech from background noise.

Keywords: Cochlear implants, speech-in-noise, speech unmasking, electroencephalography (EEG), electric acoustic stimulation (EAS)

1. Introduction

One of the biggest complaints by cochlear implant (CI) users is their difficulty to understand speech in noisy situations (Kochkin, 2005). Because speech-in-noise understanding requires multiple sensory and cognitive neural processes, there can be various causes for the difficulties experienced by CI users, which poses a significant barrier to clinical assessment and treatment.

Prior works in normal hearing (NH) listeners have identified several central processes that constitute successful speech-in-noise perception by unmasking speech from background noise. Those processes include: 1) accurate encoding of acoustic features of sound sources such as pitch (Lad et al., 2020), 2) abstraction of the target from crowded auditory scenes (Holmes & Griffiths, 2019), and 3) neural enhancement of target speech and inhibition of background noise (Choi et al., 2014; Kim et al., 2021). The above processes may cascade, meaning that the failure of acoustic feature encoding may degrade downward processes of target speech abstraction, neural enhancement of target speech and inhibition of background noise. Indeed, CI listening, which lacks the delivery of low-frequency for pitch perception, often exhibits poor performance in auditory stream segregation (Oxenham, 2008).

In the recent decade, hearing remediation through CIs has moved toward combining electric and acoustic stimulation (EAS) (Gantz & Turner, 2004; von Ilberg et al., 1999). As expected, EAS listeners with residual low-frequency acoustic hearing demonstrate improved encoding of acoustic features such as pitch (Brockmeier et al., 2010; Gantz et al., 2005; McDermott et al., 2009; Woodson et al., 2010). It has been also demonstrated that CI users with EAS perform better at speech-in-noise (SiN) tasks than electric-only (herein referred to as the “E-only”) CI users (Gantz et al., 2006; Gantz & Turner, 2004; Incerti et al., 2013; Rader et al., 2013; Turner et al., 2008). In the studies that compared EAS listening to E-only listening in the same subjects, the EAS condition exhibited significantly lower speech reception thresholds than the E-only condition (Imsiecke et al., 2020a).

However, the neural mechanism underlying improved SiN performance in EAS listeners has been unclear. Specifically, it is unclear whether better SiN performance in EAS listeners is due to the better central processes for unmasking. This study aimed to directly compare neural processes of speech unmasking between EAS and E-only listeners.

As described above, the fidelity of acoustic feature encoding may affect the separation of target speech from background noise. Supporting this idea, previous electro-physiological studies in NH listeners have correlated the fidelity of supra-threshold acoustic cue coding to SiN understanding (Anderson & Kraus, 2010; Anderson & Parbery-Clark, 2013; Hornickel et al., 2009; Song et al., 2011). Also focusing on the contribution of acoustic feature encoding to SiN understanding, previous CI studies showed that the cortical acoustical change complex (McGuire et al., 2021) and N1-P2 cortical auditory event-related potential (Berger et al., 2021) predict SiN performance.

Endogenous selective attention also contributes to the speech unmasking process. It has been reported that reaction time during a selective attention task predicts SiN performance in NH listeners (Strait and Kraus, 2011). Similarly, previous CI studies suggest that the strength of attentional modulation on cortical evoked responses to competing sounds predicts SiN ability (Lee et al., 2021; Nogueira & Dolhopiatenko, 2022).

A recent study in NH listeners suggested an electroencephalographic (EEG) index of overall speech unmasking function that combines these bottom-up and top-down neural processing, referred to as internal SNR (Kim et al., 2021). Internal SNR is the amplitude ratio of cortical evoked responses to speech (i.e., a single word) and background noise. Higher internal SNR indicates better neural unmasking of target speech from background noise. The internal SNR showed a significant positive correlation with SiN performance in NH listeners. Interestingly, better NH performers in the SiN task exhibited greater internal SNR due to their weaker cortical evoked response to background noise, not due to their stronger evoked response to the target speech, indicating that inhibiting the neural representation of noise is a driving factor of better speech unmasking.

This study aimed to compare internal SNR between EAS and E-only CI users. Findings from this study may reveal how cortical process for unmasking speech from background noise is different between EAS and E-only listening, which has not been reported by previous studies.

2. Material and Methods

2.1. Participants

Twenty-nine EAS CI users between 29 and 82 years (mean = 64.3 years, SD = 13.7 years, median = 68 years, 13 (44.8%) female) and the same number of E-only CI users (i.e., fifty eight in total) between 33 and 82 years (mean = 58.4 years, SD = 11.1 years, median = 59.3 years, 16 (55.2%) female) were recruited from the University of Iowa Cochlear Implant Research Center.

All participants had post-lingual speech impairment and were neurologically normal. The median age was 68 years for EAS users and 59 years for E-only users, exhibiting a significant difference (two-sided Wilcoxon ranksum test, p = 0.028, z-score = 2.2). The age of the first CI activation was also larger in EAS than E-only CI users (median 64 vs. 50 years, Wilcoxon ranksum test p = 0.0055, z-score = 2.8). The median duration of hearing loss before cochlear implantation was 20.0 years for EAS and 10.5 years for E-only users, but this difference was not significant between groups (p = 0.36). The median duration of device experience was significantly shorter in EAS users (22 vs. 50 months, p = 0.040, z-score = −2.1). The median threshold for unaided low frequency (i.e., 250 and 500 Hz) residual hearing of the better ear was significantly lower in EAS users (50.6 dB HL vs. 78.8 dB HL, p < 0.001, z-score = −3.7). Figure 1 shows the comparisons of age, age of implantation, duration of hearing loss, CI experience, and residual acoustic hearing threshold. The duration of hearing loss was determined based on the self-report of hearing history.

Figure 1.

Figure 1.

Open in a new tab

Comparison of demographic and audiometric factors. Red bars represent the histogram of EAS users while blue bars depict E-only users. Acoustic thresholds were averaged across 250 and 500 Hz. All the comparisons except the duration of hearing loss exhibited a significant difference (Wilcoxon ranksum test, * : p < 0.05, ** : p < 0.01, *** : p < 0.001, N.S.: p > 0.05).

Four (out of 29) E-only CI users were bilateral CI users. Of the remaining 25 E-only CI users, 12 subjects (48.0 %) had a CI in the right ear. All the 29 EAS users were unilateral implanted. 23 out of the 29 EAS users (79.3 %) had CI in the right ear. American English was the primary language for all the participants in both groups. Details of all participants (e.g., demographics, CI device types) can be found in Table S1. Duration of CI experience was obtained from clinical records.

2.2. Task design and procedures

We designed a simultaneous SiN EEG paradigm in a relatively short experimental session (15 min). All study procedures were reviewed and approved by the local Institutional Review Board.

The SiN test was implemented in custom-written Matlab scripts (R2016b, Mathworks) using the Psychtoolbox 3 toolbox (Brainard, 1997; Pelli, 1997). The test (and the simultaneous EEG recording) was conducted in an acoustically treated, electrically shielded booth with a single loudspeaker (model LOFT40, JBL) positioned at 0° azimuth angle at a distance of 1.2 m. Visual stimuli were presented via a computer monitor located 0.5 m in front of the subject at eye level. Sound levels were the same across subjects.

The trial structure was identical to the task described in (Kim et al., 2021). To summarize, while subjects fixed their gaze on the cross on the computer screen, an eight-talker babble (herein “noise”) began and continued for 2 seconds. Target words were presented 1 second after noise onset. See Figure 2A for an example auditory stimulus. Target words were always presented at 70 dB SPL. All the subjects confirmed that stimuli are clearly audible. Target words consisted of 100 monosyllabic consonant-vowel-consonant words selected from the California Consonant Test (Owens & Schubert, 1977), spoken by an American male speaker with a Midwest accent. The noise was varied randomly between two levels to yield either +7 dB (low signal-to-noise ratio; SNR) or +13 dB (high SNR). Each SNR condition presented 50 trials made with 50 uniquely different target words. These two conditions were randomly intermixed in a single block. The +7 dB SNR was found from a pilot test (i.e., done in the same sound presentation environment and materials with various different SNRs in 16 CI users who are not included in the data of the present study) to determine the SNR that yields ~50% correct mean performance that may maximize the room for across-subject variance. Thus, our EEG analysis focused on data from the +7 dB SNR condition. The +13 dB SNR condition was included as an easier condition to validate that poorer SNR degrades SiN performance. 100 ms after acoustic stimuli, participants were presented with four possible words printed in the center of the screen, and provided their answer by pressing a number key that corresponds to the target word they thought they had heard. No feedback was given regarding accuracy.

Figure 2.

Figure 2.

Open in a new tab

A. An example waveform of stimuli. Gray: Noise, Green: Target word. B. Grand-average evoked responses measured in global field power (GFP) from +7dB SNR condition. The vertical lines around the peak error bar depict standard errors across subjects. Two different peaks from noise onset period (0.05 – 0.25 seconds) and target onset period (1.05 – 1.25 seconds) were detected and used for calculation of the internal SNR. Red: EAS users, Blue: E-only users. C. Comparison of internal SNRs between the groups of EAS (red) and E-only (blue) users. The three horizontal lines in each box represent 75 percentile, median, and 25 percentiles. Filled circles represent individual subjects’ internal SNR values. Asterisk (*) indicates a statistically significant difference between groups (two-sample t-test, p = 0.019).

2.3. EEG acquisition and preprocessing

Electrical activity (EEG) was recorded on the scalp during the SiN task at a sampling rate of 2048 Hz using the BioSemi ActiveTwo system with 64 electrode system arranged in a 10–20 placement system. Electrodes that coincide with the CI magnet position were excluded. Remaining number of electrodes (after removing those missing ones as well as bad electrodes with electric noises) ranged from 42 to 61 (mean 55.0, standard deviation 4.14) for EAS users and 51 to 56 (mean 53.8, standard deviation 1.58) for E-only users. There was no significant difference in the number of remaining electrodes between the groups (two-sample t-test, p = 0.15). Physiological data and stimulus timing triggers were recorded in ActiView (BioSemi). Data from each channel were filtered offline between 1 to 50 Hz using a 2048-point zero-phase non-causal FIR filter. Epochs were extracted from −500 ms to 2.5 seconds associated with the onset of auditory stimuli (i.e., the babble noise). After baseline calibration using an average voltage of −200 to 0 ms, the epochs were resampled to 256 Hz.

Next, electric field-evoked artifacts introduced by CIs and eye-blink artifacts were removed using independent component analysis (implemented in the Matlab EEGLab toolbox; (Delorme & Makeig, 2004). CI artifact and eye blink-related components were determined by visually inspecting the temporal and spatial pattern of each independent component. The following criteria were used to determine CI artifacts: 1) The temporal pattern has sharp transients time-locked to the stimulus onset and offset. 2) The temporal pattern is consistent across trials. 3) The topography includes sharp spatial transients. Out of ~60 independent components in each subject, the number of removed independent components (because contaminated by CI and eye-blink artifacts) ranged from 2 to 22. The mean number of removed components were 6.8 and 10.0 in EAS and E-only subjects, respectively. The number of removed components was significantly larger in E-only subjects than EAS subjects (two-sample t-test, p < 0.001). Eyeblink artifacts were determined by visual inspection.

Then epochs with maximum amplitude exceeding 100 μV were discarded from the following analyses. After this preprocessing, the number of remaining artifact-free epochs ranged from 31 to 49 in each condition in each subject (i.e., reduced by 2 % – 28 % from original 50 trials in each condition). The mean number of epochs were 42.6 and 43.0 for EAS subjects (for +13 dB and +7 dB SNR conditions, respectively), and 45.0 and 44.4 for E-only subjects (for +13 dB and +7 dB SNR conditions, respectively). There was no significant difference in the number of epochs between groups (EAS vs. E-only, two-sample t-test, p = 0.22).

2.4. EEG indexing: Peak finding and “internal SNR”

Internal SNR (Kim et al., 2021) was computed for each subject by taking the peak-amplitude ratio of evoked responses to speech and noise in the low (+7 dB) SNR condition. To compute internal SNRs, the amplitudes of evoked response to noise and speech were found using the following steps:

  1. The artifact-free epochs were averaged for each subject in each SNR condition to yield evoked response time courses. The dimensionality of evoked responses from multiple electrodes (please see the number of remaining electrodes described above) was reduced by taking the spatial standard deviation of each time point, or global field power [GFP: “A reference-independent descriptor of the potential field,” (Skrandies, 1990)] for each subject. The GFP time courses were smoothed by a zero-phase FIR low pass filter at 8 Hz.

  2. Then we applied the Jackknife approach (Miller et al., 1998, 2009), which took the following steps iii – v.

  3. Compute N grand average of {N-1} subjects by taking one subject out each time, where N = the number of subjects in each group (i.e., 29).

  4. For each jackknife grand average (across {N-1} subjects), the peak was found as the local maximum within the pre-defined time range of typical cortical auditory evoked potentials (i.e., 50 – 250 ms) after the onset of background noise (0 s) and target speech (1 s).

  5. Then the individual peak values were used for later statistical analyses (e.g., two-sample t-test of peak amplitudes between groups). Mathematically, this jackknife process diminishes the standard deviation across subjects by the factor of {N-1} (i.e., 28 in our case). For its adjustment, {N-1} was multiplied to standard deviation terms in later statistical analyses (e.g., two-sample t-test of peak amplitudes between groups). The same adjustment was applied to the comparison and visualization of internal SNR (Figure 2C).

2.5. Statistical analysis

A two-sample t-test was applied to compare internal SNRs between the groups of EAS and E-only users.

3. Results

3.1. Speech-in-noise performance

SiN performance (i.e., percent correct responses) ranged from 20 % to 88 % in the +13 dB SNR condition (mean 64.3 %, standard deviation 15.1 %) and ranged from 30 % to 82 % in the +7 dB SNR condition (mean 54.1 %, standard deviation 12.0 %) for all 58 participants. As expected, SiN performance was significantly poorer in the main experimental condition (i.e., +7 dB SNR) than the +13 dB SNR condition (paired t-test, p<0.001, t-score = −7.3).

In the +7 dB SNR condition, SiN performance ranged from 30 % to 68 % (mean 51.7 %, standard deviation 9.8 %) in the EAS users and ranged from 30 % to 82 % (mean 56.6 %, standard deviation 13.7 %) in the E-only users. No significant difference was found between the groups (two-sample t-test, p = 0.13). In the +13 dB SNR condition, EAS users’ SiN performance ranged from 20 % to 82 % (mean 64.1 %, standard deviation 14.7 %) and E-only users’ performance ranged from 26 % to 88 % (mean 65.4, standard deviation 15.7 %). No significance difference between performance was found in the 13 dB SNR condition either (two-sample t-test, p = 0.56)

3.2. Internal SNR: +7 dB SNR condition

Figure 2B shows grand-average evoked responses to illustrate how peak amplitudes of evoked potential following noise and target speech onsets were found. The position of each vertical line indicates the position of each subject’s GFP peak, while its length indicates the standard error of GFP across subjects at that time point. Internal SNR was computed for each subject by dividing the GFP peak amplitude following target speech onset with the GFP peak amplitude following the noise onset. Figure 3C compares internal SNRs, or the ratios of evoked potential to speech and noise, between the groups of EAS and E-only users. EAS users exhibited significantly larger internal SNRs than E-only users (two-sample t-test, p = 0.019, t-score = 2.42).

Figure 3.

Figure 3.

Open in a new tab

A. Grand average GFP time courses before (dashed curves) and after (solid curves) artifact removal. B. Grand average topographies at the GFP peak positions following the noise onset. C. Grand average topographies at the GFP peak positions following the speech onset (after artifact removal).

Neither noise-evoked nor target speech-evoked potentials showed significant difference between groups (two-sample t-test, p = 0.22, t = −1.25 at the noise onset, p = 0.68, t = 0.41 at the speech onset).

3.3. Comparisons of cochlear implant EEG artifacts

CI artifact could confound the internal SNR analysis. To investigate the effect of CI artifact on the above EEG results, we did three further analyses described below.

First, we compared GFP peak amplitudes between E-only and EAS groups’ EEG data before passing the data through our ICA-based artifact removal process. The GFP time courses before artifact removal are shown as dashed curves in Figure 3A. E-only users’ peak artifact amplitude (in GFP) was bigger than EAS users’ (two-sample t-test, t = −2.50, p = 0.016), which raises a possibility that CI artifact confounded the evoked response comparison between the groups.

Second, we made a 2-by-2 factorial comparison of the scalp topographies of peak voltages 1) between before- and after the ICA-based artifact removal and 2) between E-only and EAS data. The topographies are shown in Figure 3B. No similarity is evident between the topographies before and after artifact removal.

Third, we performed a correlation analysis between the peak GFP amplitudes before and after artifact removal. No significant correlation was found; Pearson correlation coefficients between the log amplitude of artifact and the peak amplitude after artifact removal was −0.28 (p = 0.15) for EAS users (red circles in Figure 4), −0.31 (p = 0.096) for E-only users (blue circles in Figure 4), and −0.23 (p = 0.078) for all the participants.

Figure 4.

Figure 4.

Open in a new tab

Relationship between the peak GFP amplitudes before and after artifact removal.

3.4. Results from +13 dB SNR condition

As stated earlier, our EEG analysis focused on the +7 dB SNR condition because it was determined as the condition that yields the greatest room for variance across subjects. +13 dB SNR condition (where the noise level was 6 dB lower than the +7 dB SNR condition) was tested just to validate the noise level effect on the performance, However, the same analysis was applied to data from +13 dB SNR condition for a comparison.

Figure 5 shows the stimulus structure (A), grand-average evoked responses in GFP (B), and the comparison of internal SNRs between EAS and E-only subjects (C) computed from the +13 dB SNR condition data. This condition did not yield a significant difference of internal SNR between groups (two sample t-test for EAS vs. E-only, T = −0.26, p = 0.80).

Figure 5.

Figure 5.

Open in a new tab

A. An example waveform of stimuli. Gray: Noise, Green: Target word. B. Grand-average evoked responses measured in global field power (GFP) from +13 dB SNR condition. C. Comparison of internal SNRs between the groups of EAS (red) and E-only (blue) users.

4. Discussion

We compared neural activities underlying SiN perception between EAS and E-only CI users by recording scalp EEG during a SiN task and taking the amplitude ratio of evoked responses to speech and noise, or internal SNR (Kim et al., 2021). EAS users showed greater internal SNR than E-only users, which suggests that EAS users’ cortical processing unmasks speech from background noise more effectively.

What factors underlie internal SNR differences?

Internal SNR has been suggested as an index that compares the evoked potential to speech vs. the neural response to noise. It has been reported that internal SNR predicts SiN performance (Kim et al., 2021) and how tolerable an individual is to background noise during a SiN task (Kim et al., 2022). An individual’s internal SNR can be larger than others because of stronger response to target speech or weaker response to noise, or combination of those two. In the present study, greater internal SNR in EAS users was achieved by the combination of their neural responses to target speech and noise, but not by either neural response alone. However, the neural response to noise made a greater contribution (t-score −1.25 at the noise onset vs. 0.41 at the speech onset). Interestingly, this is consistent with our previous finding in NH listeners that good SiN performers exhibited weaker neural response to noise than poor SiN performers (Kim et al., 2021).

A possible neural mechanism that explains the weaker neural response to background noise is “sensory gain control (Hillyard et al., 1998)” or “adaptive filtering (Giard et al., 2000)” during auditory selective attention that may occur in superior temporal plane for early (100–200ms) cortical evoked responses (Choi et al., 2013; Hillyard et al., 1973; Mesgarani & Chang, 2012). The latency and topographies of GFP peaks that contributed to the internal SNR computation in the present study support this interpretation that selective attention played a role. This interpretation is consistent with a recent study that reported a relationship between selective attention and CI outcomes (Nogueira & Dolhopiatenko, 2022).

The difference in the attentional modulation of auditory cortical responses may originate from a few different factors. It may be due to the preserved low-frequency acoustic hearing that provides effective acoustic cues for auditory grouping, e.g., pitch (Darwin, 1997; Oxenham, 2008). Since we used fixed timing of events (i.e., noise and target word onsets) in our experiment, predicted timing could also contribute through neural phase locking (Arnal & Giraud, 2012).

Alternative interpretations.

The evoked response to target speech in our analysis is similar to the acoustic change complex (McGuire et al., 2021). In that sense, an alternative interpretation of the internal SNR result is that EAS users exhibited a greater sensitivity to acoustical change when speech acoustics arose from the acoustics of background noise, like reported in (Faucette & Stuart, 2017). Although this interpretation does not explain the difference in neural response to noise, this sensitivity difference could still contribute to internal SNR since it is the combination of both neural responses (to noise and speech).

Did CI artifact confound the result?

Significantly larger CI artifact in E-only users may confounded the internal SNR result. However, we cautiously claim that the CI artifact effect was negligible because of the following reasons. First, the scalp distribution of voltage (topography) at the GFP peak (Figure 3B) was different before vs. after artifact removal, although it was rather similar between groups after artifact removal. Second, there was no significant correlation between the peak amplitude before and after artifact removal (Figure 4).

However, it is impossible to completely rule of the possibility of residual CI artifact leaking into the analysis of evoked neural responses. A future study may combine an artifact-free neuroimaging modality for CI users such as positron emission tomography.

Limitations of the current study

Unlike the EEG data, our behavioral data (i.e., accuracy) did not exhibit a significant difference between the groups. This inconsistency can be due to the following reasons. First, the differences in demographic factors between groups. As described in the Participants section (2.1), our EAS subjects were older, implanted at an older age, and less experienced with CI than our E-only subjects. All these demographic differences were disadvantageous for our EAS subjects’ speech performance (Mosnier et al., 2014). It is possible that, despite poorer cortical processing for speech unmasking, our E-only subjects did not perform worse due to longer device experience started from their younger age. However, the fact that demographic factors did not match between the groups is a limitation of our study. A future study will seek better control of demographic factors between groups. Second, a few previous studies [e.g., (Imsiecke et al., 2020b)] compared performance of E-only vs. EAS conditions in the same subjects. Although such within-subject comparison could be advantageous for the control of CI artifact, we avoided such comparison to prevent the input bandwidth difference confounding the result.

Our choice of speech material (i.e., California Consonant Test) could impact the result and may not be generalized to different SiN tasks. Indeed, Reiss et al. reported that some E-only users perform consonant discrimination unexpectedly well (Reiss et al., 2008).

Our high-SNR condition (+13 dB SNR) did not show consistent results. Although it was not our pre-determined SNR to investigate outcome variances, the inconsistency in the result is a limitation of this study. A potential interpretation is that SiN performance with lower noise level is not limited by how well listeners suppress the noise. A future study should focus on the SNR effect on cortical processes for SiN understanding.

Implications.

Our results have both theoretical and clinical implications. Theoretically, our internal SNR method revealed a critical neural subsystem involved during SiN processing and the importance of acoustic cues for speech unmasking. Clinically, our results suggest that a relatively short (~15 minutes) SiN-EEG paradigm can assess this crucial neural process for SiN understanding in CI listeners.

In the present paper, we exhibited group-level differences of evoked responses only. Although our results demonstrated how this simple EEG analysis successfully reveals differences in neural activities, future studies may pursue further understanding of SiN mechanisms by adopting individual differences approach (e.g., (Bonnard et al., 2018) and/or extended EEG analyses such as induced oscillations.

5. Conclusion

This study found that the amplitude ratio of evoked responses to speech and noise, or internal SNR, was significantly larger in the EAS than E-only users. This result may indicate that CI users with a greater degree of residual acoustic hearing exhibit enhanced unmasking of speech from background noise during a SiN task.

Supplementary Material

1

Highlights:

  • EAS users exhibit better cortical response to speech-in-noise than E-only users.

  • EAS users did not perform poorer than E-only users with longer CI experiences.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Anderson S, & Kraus N. (2010). Objective Neural Indices of Speech-in-Noise Perception. Trends in Amplification, 14(2), 73–83. 10.1177/1084713810380227 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson S, & Parbery-Clark A. (2013). Auditory brainstem response to complex sounds predicts self-reported speech-in-noise performance. Journal of Speech, Language, and Hearing Research, 56(February), 31–44. 10.1044/1092-4388(2012/120043)in [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arnal LH, & Giraud A-L (2012). Cortical oscillations and sensory predictions. Trends in Cognitive Sciences, 16(7), 390–398. 10.1016/j.tics.2012.05.003 [DOI] [PubMed] [Google Scholar]
  4. Berger JI, Gander PE, Kim S, Schwalje AT, Woo J, Na Y, Holmes A, Hong J, Dunn C, Hansen M, Gantz B, McMurray B, Griffiths TD, & Choi I. (2021). Neural correlates of individual differences in speech-in-noise performance in a large cohort of cochlear implant users. BioRxiv, 2021.04.22.440998. 10.1101/2021.04.22.440998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bonnard D, Schwalje A, Gantz B, & Choi I. (2018). Electric and acoustic harmonic integration predicts speech-in-noise performance in hybrid cochlear implant users. Hearing Research, 367, 223–230. 10.1016/j.heares.2018.06.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brockmeier SJ, Peterreins M, Lorens A, Vermeire K, Helbig S, Anderson I, Skarzynski H, van de Heyning P, Gstoettner W, & Kiefer J. (2010). Music Perception in Electric Acoustic Stimulation Users as Assessed by the Mu.S.I.C. Test. Cochlear Implants and Hearing Preservation, 67, 70–80. 10.1159/000262598 [DOI] [PubMed] [Google Scholar]
  7. Choi I, Rajaram S, Varghese LA, & Shinn-Cunningham BG (2013). Quantifying attentional modulation of auditory-evoked cortical responses from single-trial electroencephalography. Frontiers in Human Neuroscience, 7(April), 1–19. 10.3389/fnhum.2013.00115 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Choi I, Wang L, Bharadwaj H, & Shinn-Cunningham B. (2014). Individual differences in attentional modulation of cortical responses correlate with selective attention performance. Hearing Research, 314, 10–19. 10.1016/j.heares.2014.04.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Darwin CJ (1997). Auditory grouping. In Trends in Cognitive Sciences (Vol. 1, Issue 9, pp. 327–333). Elsevier Ltd. 10.1016/S1364-6613(97)01097-8 [DOI] [PubMed] [Google Scholar]
  10. Delorme A, & Makeig S. (2004). EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. Journal of Neuroscience Methods, 134(1), 9–21. 10.1016/J.JNEUMETH.2003.10.009 [DOI] [PubMed] [Google Scholar]
  11. Faucette SP, & Stuart A. (2017). Evidence of a speech evoked electrophysiological release from masking in noisea). The Journal of the Acoustical Society of America, 142(2), EL218. 10.1121/1.4998151 [DOI] [PubMed] [Google Scholar]
  12. Gantz BJ, & Turner C. (2004). Combining acoustic and electrical speech processing: Iowa/Nucleus hybrid implant. Acta Oto-Laryngologica, 124(4), 344–347. 10.1080/00016480410016423 [DOI] [PubMed] [Google Scholar]
  13. Gantz BJ, Turner C, & Gfeller KE (2006). Acoustic plus Electric Speech Processing: Preliminary Results of a Multicenter Clinical Trial of the Iowa/Nucleus Hybrid Implant. Audiology and Neurotology, 11(Suppl. 1), 63–68. 10.1159/000095616 [DOI] [PubMed] [Google Scholar]
  14. Gantz BJ, Turner C, Gfeller KE, & Lowder MW (2005). Preservation of Hearing in Cochlear Implant Surgery: Advantages of Combined Electrical and Acoustical Speech Processing. The Laryngoscope, 115(5), 796–802. 10.1097/01.MLG.0000157695.07536.D2 [DOI] [PubMed] [Google Scholar]
  15. Giard M-H, Fort A, Mouchetant-Rostaing Y, & Pernier J. (2000). NEUROPHYSIOLOGICAL MECHANISMS OF AUDITORY SELECTIVE ATTENTION IN HUMANS. Frontiers in Bioscience, 5, 84–94. [DOI] [PubMed] [Google Scholar]
  16. Hillyard SA, Hink RF, Schwent VL, & Picton TW (1973). Electrical signs of selective attention in the human brain. Science, 182(4108), 177–180. [DOI] [PubMed] [Google Scholar]
  17. Hillyard SA, Vogel EK, & Luck SJ (1998). Sensory gain control (amplification) as a mechanism of selective attention: electrophysiological and neuroimaging evidence. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 353(1373), 1257–1270. 10.1098/rstb.1998.0281 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Holmes E, & Griffiths TD (2019). ‘Normal’ hearing thresholds and fundamental auditory grouping processes predict difficulties with speech-in-noise perception. Scientific Reports, 9(1), 1–11. 10.1038/s41598-019-53353-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hornickel J, Skoe E, Nicol T, Zecker S, & Kraus N. (2009). Subcortical differentiation of stop consonants relates to reading and speech-in-noise perception. Proceedings of the National Academy of Sciences of the United States of America, 106(31), 13022–13027. 10.1073/pnas.0901123106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Imsiecke M, Krüger B, Büchner A, Lenarz T, & Nogueira W. (2020a). Interaction Between Electric and Acoustic Stimulation Influences Speech Perception in Ipsilateral EAS Users. Ear and Hearing, 41(4), 868. 10.1097/AUD.0000000000000807 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Imsiecke M, Krüger B, Büchner A, Lenarz T, & Nogueira W. (2020b). Interaction Between Electric and Acoustic Stimulation Influences Speech Perception in Ipsilateral EAS Users. Ear and Hearing, 41(4), 868. 10.1097/AUD.0000000000000807 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Incerti P. v., Ching TYC, & Cowan R. (2013). A systematic review of electric-acoustic stimulation: Device fitting ranges, outcomes, and clinical fitting practices. Trends in Amplification, 17(1), 3–26. 10.1177/1084713813480857 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kim S, Schwalje AT, Liu AS, Gander PE, McMurray B, Griffiths TD, & Choi I. (2021). Pre- and post-target cortical processes predict speech-in-noise performance. NeuroImage, 228, 117699. 10.1016/j.neuroimage.2020.117699 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kim S, Wu Y-H, Bharadwaj HM, & Choi I. (2022). Effect of Noise Reduction on Cortical Speech-in-Noise Processing and Its Variance due to Individual Noise Tolerance. Ear & Hearing, 43(3), 849–861. 10.1097/AUD.0000000000001144 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kochkin S. (2005). MarkeTrak VII: Customer satisfaction with hearing instruments in the digital age. Hearing Journal, 58(9), 30–43. 10.1097/01.HJ.0000286545.33961.E7 [DOI] [Google Scholar]
  26. Lad M, Holmes E, Chu A, & Griffiths TD (2020). Speech-in-noise detection is related to auditory working memory precision for frequency. Scientific Reports 2020 10:1, 10(1), 1–8. 10.1038/s41598-020-70952-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lee JH, Orr M, Shim H, & Choi I. (2021). Auditory selective attention and speech-in-noise in cochlear implants. The Journal of the Acoustical Society of America, 150(4), A339. 10.1121/10.0008509 [DOI] [Google Scholar]
  28. McDermott H, Sucher C, & Simpson A. (2009). Electro-Acoustic Stimulation. Audiology and Neurotology, 14(Suppl. 1), 2–7. 10.1159/000206489 [DOI] [PubMed] [Google Scholar]
  29. McGuire K, Firestone GM, Zhang N, & Zhang F. (2021). The Acoustic Change Complex in Response to Frequency Changes and Its Correlation to Cochlear Implant Speech Outcomes. Frontiers in Human Neuroscience, 15, 629. 10.3389/FNHUM.2021.757254/BIBTEX [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mesgarani N, & Chang EF (2012). Selective cortical representation of attended speaker in multi-talker speech perception. Nature, 485(7397), 233–236. 10.1038/nature11020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Miller J, Patterson T, & Ulrich R. (1998). Jackknife-based method for measuring LRP onset latency differences. Psychophysiology, 35(1), 99–115. 10.1017/S0048577298000857 [DOI] [PubMed] [Google Scholar]
  32. Miller J, Ulrich R, & Schwarz W. (2009). Why jackknifing yields good latency estimates. Psychophysiology, 46(2), 300–312. 10.1111/J.1469-8986.2008.00761.X [DOI] [PubMed] [Google Scholar]
  33. Mosnier I, Bebear JP, Marx M, Fraysse B, Truy E, Lina-Granade G, Mondain M, Sterkers-Artières F, Bordure P, Robier A, Godey B, Meyer B, Frachet B, Poncet C, Bouccara D, & Sterkers O. (2014). Predictive Factors of Cochlear Implant Outcomes in the Elderly. Audiology and Neurotology, 19(Suppl. 1), 15–20. 10.1159/000371599 [DOI] [PubMed] [Google Scholar]
  34. Nogueira W, & Dolhopiatenko H. (2022). Predicting speech intelligibility from a selective attention decoding paradigm in cochlear implant users. Journal of Neural Engineering, 19(2), 026037. 10.1088/1741-2552/AC599F [DOI] [PubMed] [Google Scholar]
  35. Owens E, & Schubert ED (1977). Development of the California Consonant Test. Journal of Speech, Language, and Hearing Research, 20, 463–474. 10.3102/00028312040003655 [DOI] [PubMed] [Google Scholar]
  36. Oxenham AJ (2008). Pitch Perception and Auditory Stream Segregation: Implications for Hearing Loss and Cochlear Implants. Trends in Amplification, 12(4), 316–331. 10.1177/1084713808325881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rader T, Fastl H, & Baumann U. (2013). Speech perception with combined electricacoustic stimulation and bilateral cochlear implants in a multisource noise field. Ear and Hearing, 34(3), 324–332. 10.1097/AUD.0B013E318272F189 [DOI] [PubMed] [Google Scholar]
  38. Reiss LAJ, Gantz BJ, & Turner CW (2008). Cochlear Implant Speech Processor Frequency Allocations May Influence Pitch Perception. Otology & Neurotology : Official Publication of the American Otological Society, American Neurotology Society [and] European Academy of Otology and Neurotology, 29(2), 160. 10.1097/MAO.0B013E31815AEDF4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Skrandies W. (1990). Data Reduction of Multichannel Fields: Global Field Power and Principal Component Analysis. Brain Topography, 3(1). [DOI] [PubMed] [Google Scholar]
  40. Song JH, Skoe E, Banai K, & Kraus N. (2011). Perception of speech in noise: Neural correlates. Journal of Cognitive Neuroscience, 23(9), 2268–2279. 10.1162/jocn.2010.21556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Turner C, Gantz BJ, & Reiss L. (2008). Integration of acoustic and electrical hearing. Journal of Rehabilitation Research and Development, 45(5), 769–778. 10.1682/JRRD.2007.05.0065 [DOI] [PubMed] [Google Scholar]
  42. von Ilberg C, Kiefer J, Tillein J, Pfenningdorff T, Hartmann R, Stürzebecher E, & Klinke R. (1999). Electric-acoustic stimulation of the auditory system. New technology for severe hearing loss. ORL; Journal for Oto-Rhino-Laryngology and Its Related Specialties, 61(6), 334–340. 10.1159/000027695 [DOI] [PubMed] [Google Scholar]
  43. Woodson EA, Reiss LAJ, Turner CW, Gfeller K, & Gantz BJ (2010). The Hybrid Cochlear Implant: A Review. Adv Otorhinolaryngol. Basel, Karger, 67, 125–134. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1854398-supplement-1.xlsx (16.3KB, xlsx)

RESOURCES