Goals: The goal of this line of research is to better understand auditory neural encoding of electrical stimulation and its relationship to perceptual outcomes in cochlear implant (CI) users.
Material and Methods: This presentation summarizes results from three sets of studies. The first identified cochlear nerve (CN) response properties that predict speech perception outcomes. The second developed distinct electrophysiological measures to separately assess the number of available CN fibers and their neural health. The third examined relationships among CN neural synchrony, perceptual sensitivity to temporal envelope cues, and speech perception outcomes.
Results: Both the quality of the electrode-to-neuron interface (ENI) and the speed of recovery from neural adaptation were significant predictors of speech perception performance. Poorer ENIs and slower adaptation recovery were associated with poorer speech perception, particularly in noise. Adaptation recovery explained additional unique variance beyond ENI measures alone, and together these factors accounted for substantially more variance under high-noise conditions, explaining approximately 50% of the variance in speech recognition scores at a +5 dB signal-to-noise ratio.
Sensitivity to an increase in the interphase gap (IPG) of biphasic pulses provided an estimate of CN fiber count, whereas neural synchrony primarily reflected the health of existing CN fibers. Stimulation level offsets and the IPG effect on the maximum slope of the eCAP input/output function were effective for comparing CN fiber counts across groups. In contrast, the phase-locking value (PLV) was identified as the preferred metric for assessing neural synchrony. Reduced neural synchrony was further associated with prolonged within-channel gap detection thresholds and poorer speech perception.
Conclusions: Speech perception outcomes in CI users is strongly influenced by both ENI quality and intrinsic CN response properties. Distinct electrophysiological metrics can dissociate CN fiber quantity from neural health, with IPG-based measures indexing fiber count and PLV indexing neural synchrony. Reduced neural synchrony is linked to impaired temporal processing and poorer speech perception, particularly in challenging listening conditions. Together, these findings emphasize the importance of using multiple, biologically grounded neural metrics to predict perceptual outcomes and support individualized CI assessment and o

Goals
Previously, we derived an electrophysiological marker for neural health profiles in the animal model (Konerding et al. 2025a). This electrically-evoked compound-action potential (eCAP)-derived Failure Index (FI) was indicating the presence of a mechanically induced, spatially restricted lesion of the spiral ganglion neurons. Thereby, a high FI indicated the failure to effectively transmit current into neural signals and the size of the elevation indicated the presence, site and size of a lesion. Here, we translated the FI to clinical recordings of MED-EL CI users.
Material and Methods
For the retrospective study (Konerding et al. 2025b), we selected patient data from the database of the German Hearing Center Hannover recorded 2017 to 2024. We included 199 post-lingually and 79 pre-lingually deafened ears. The eCAPs were recorded at the Deutsche Hörzentrum Hannover between August 2017 and August 2024. To accommodate for differences in stimulation-phase duration, we calculated the FI as the charge needed to reach saturation level, divided by the eCAP amplitude reached at this maximal charge level. Changes in FI were analysed with regard to factors expected to impact cochlear health, such as age, duration/cause of deafness, and stimulation position (apical to basal). As proxy of cochlear health, we used speech perception in quiet and noise and correlated with eCAP-derived measures.
Results
Averaged FIs over all contacts of a CI were stable within the analysis period (3rd month to 1st year postoperatively). The averaged FI increased with age and was elevated for etiologies associated with higher SGN loss. Utilizing 3D information from cone beam-computed tomography scans, we confirmed that the FI was independent of distance to the modiolus. The FI showed individual patterns along the array with maxima usually at basal contacts, corresponding to reduced cochlear health at high-frequency regions. In a selected group of post-lingually deaf ears, we confirmed significant correlation with speech perception, with higher goodness of fit than for other eCAP measures (i.e., threshold, slope and amplitude).
Conclusion
The results are in line with the hypothesis that the FI may serve as clinical tool to identify CI-implanted ears with reduced cochlear health and contacts close to areas of reduced SGN survival and/or integrity.

The electrically evoked compound action potential (eCAP) reflects the combined responsiveness to electrical stimulation of the spiral ganglion neurons (SGNs) that form the auditory nerve. In animal studies eCAP measures correlate strongly with SGN survival, suggesting that these neural health measures might be used clinically to assess neural health status, or even predict cochlear implant (CI) outcome. While several studies have indeed reported solid correlations between eCAP measures and speech perception (e.g., Zamaninezhad et al., 2023, Front Integr Neurosci 17), there have also been reports to the contrary. We hypothesize that major factors influencing the heterogeneity in human data include the extent, duration, and etiology of hearing loss.
In the present study we examine the potential of the eCAP, and various derivatives, as a neural health measure in a population of >180 MED-EL CI recipients. In this relatively large population, the influence of etiology on eCAP measures may also be observed. eCAPs were recorded intra-operatively with a short (2.1 µs) and a long (30 µs) inter-phase gap (IPG) using AutoART in MED-EL’s clinical MAESTRO software. Both the absolute eCAP measures (e.g., amplitude, slope, dynamic range, latency) as well as their IPG effects were evaluated. Etiology of hearing loss was categorized, and pre-operative pure tone audiograms were used to determine the extent of the hearing loss.
Etiology significantly affected multiple absolute eCAP measures and IPG effects, especially in the cochlear base. Most notably, eCAP measures in patients with auditory neuropathy were significantly different from those in patients with presynaptic pathologies. Across all groups, neural health measures scored lower in the cochlear base than in the apex. The extent of low-frequency hearing preservation moderately correlated with several eCAP measures including latency and maximum amplitude. Moderate correlations were observed between speech perception in noise and several eCAP measures, including latency and slope.
We conclude that given the myriad of causes for hearing loss, each affecting different structures of the hearing organ, we should be taking these causes into account when attempting to determine neural health objectively. Furthermore, optimized research tools may provide more relevant insights than clinically approved software.

Goals: We present a pipeline to create digital twins (DTs) of the electro-neural interface (ENI) for cochlear implant (CI) recipients that estimate the number of healthy auditory nerve fibers (ANFs) along the length of the cochlea. We propose an objective neural health score (NHS) that depends on the estimate of the count of healthy ANFs. We hypothesize NHS correlates with post-operative speech-in-noise recognition scores.

Materials and Methods: DTs were constructed for 13 CI recipients. CI stimulation was simulated by solving Poisson’s equation to estimate extracellular voltages along the ANFs using computed tomography-based tissue maps [1]. Custom tissue electrical resistivities were estimated per subject to obtain accurate predictions of intra-cochlear trans-impedances [2]. Stimulation of single ANF action potentials was simulated using the generalized Schwarz-Eikhof-Frijns model [3-5]. We estimated the number of healthy ANFs (up to ~500) in each of 75 ANF bundles along the length of the cochlea to obtain accurate simulations of compound action potential amplitude growth functions (AGFs) [2].

We propose an aggregate neural health score NHS=1/75 ∑_(i=1)^75 (log⁡(N_i/γ+ε)) , where N_i is the number of estimated healthy fibers in bundle i. NHS assigns a lower score for individuals with some ANF bundles estimated to contain fewer than roughly γ (set to 50) healthy fibers, indicating regions of poor neural health in the cochlea.

We compared NHS to unilateral CI-aided speech recognition scores on the AzBio sentence recognition tests with speech presented at 65 dB SPL and multi-talker babble presented at 60 dB SPL, tested at 6-12 months post-CI activation.

Results: Mean and standard deviation NHS and speech-in-noise scores were 0.20 +/- 0.68 and 38% +/- 20%, respectively. Pearson’s correlation coefficient between NHS and speech-in-noise scores was of moderate effect size and statistically significant (R = 0.61, p < 0.03).

Conclusions: Correlation between NHS and speech-in-noise recognition suggests that patient-specific ENI DTs capture clinically meaningful variation in neural health. These DTs can not only predict local neural health but also simulate complex stimuli, e.g., to measure refractory effects in stimulation sequences, which will be demonstrated at the conference. Future work will examine whether individualized, DT-guided CI programming can translate these neural health estimates into improved speech recognition.

Goals
Cochlear implant (CI) recipients exhibit wide variability in speech outcomes, making it challenging for clinicians to assess expected performance and personalized rehabilitation strategies. Although multiple factors contribute to outcome variability, the quality of the electrode–neural interface (ENI) is believed to be a primary underlying source. Advanced diagnostic and predictive tools are therefore needed during aftercare to support troubleshooting, guide personalized counseling, and improve patient outcomes. The objective of this work was to identify promising ENI derived biomarkers and quantify the extent to which speech outcome variability can be explained using a multivariable, data driven modeling approach.

Methods
Data from 42 adult CI recipients implanted with the Nexa™ device in feasibility studies were analysed. Threshold levels of auditory nerve eCAPs were measured with Neural Response Telemetry in monopolar and focused stimulation configurations. As eCAPs are influenced by electrode positioning, electrode–modiolar proximity was quantified from postoperative CT scans and incorporated as an anatomical descriptor of the electrode–neural interface. Different combinations of eCAP and modiolar proximity derived features were evaluated as candidate biomarkers in multivariate models. Speech perception outcomes—including QSMD, Listening in Noise Test (LIT) scores, and speech in quiet performance (CVC word scores)—measured at approximately six months post activation were used as dependent variables. Additionally, simulated data from a three dimensional electro anatomical model (3D EAM) are being explored to study relationship between electrode position, neural characteristics, and electrophysiological features.

Results
Model performance was evaluated using the coefficient of determination (R²) between predictions of the multivariate model and the clinically measured QSMD, LIT and Word scores. Preliminary results indicate multivariate models accounted for 40% & 58% of variability in the LIT & Words outcome respectively. Findings from 3D EAM simulations will be presented to provide additional insight into potential mechanistic relationships.

Conclusion
These promising results indicate that ENI-derived biomarkers explain a substantial proportion of speech outcome variability and provide clinically relevant insight into the electrode–neural interface. These models can assist clinicians towards more personalized fitting and improving CI outcomes.

Background and Rationale: Children with early-onset unilateral deafness (UD) hear from only the left or right ear, potentially disrupting normal projections from each ear to contralateral auditory cortices and hemispheric specialization for auditory processing.

Methods: This study measured cortical responses to sound in 31 children who received a unilateral cochlear implant (CI) for pre- or perilingual asymmetric hearing loss, providing a unique opportunity to assess function in both intact and deprived auditory pathways. All children were tested at initial stages of CI-use (mean (SD) CI-use = 7.16 (3.83) days) and were grouped by side of UD: 18 children (58.1%) presented with right ear hearing/left UD and 13 children (41.9%) had left ear hearing/right UD. Median (IQR) age at CI activation was 3.27 (1.83–5.10) years old, with a median (IQR) duration of auditory deprivation of 1.74 (1.02–2.61) years.

Results: There were no significant differences in demographic or hearing history factors between groups. Compared to peers with normal-hearing, children with both left and right UD showed weaker auditory cortical responses than children with normal hearing (mean [SD] = 10.4 [8.17] nAm) (p = 0.001 and p=0.004, respectively). In children with normal hearing, activity evoked from each ear generally lateralized to the contralateral AC. Children with UD rarely showed this pattern (n= 5 [29.4%] for right ear hearing, n = 2 [22.2%] for left ear hearing). Rather, there was a consistency in hemisphere dominance regardless of the stimulated ear (r = 0.53, 95% CI [0.17, 0.76], p < 0.01), with a higher likelihood of right than left AC dominance (n = 11 (35.4%) right AC, n = 6 (19.4%) left AC). The dominant hemisphere could not be predicted by hearing history or ear (left or right) of UD.

Conclusions: These findings reveal hemispheric- rather than ear-driven auditory processing in children with either left or right UD and have potential consequences for neuroplasticity, auditory processing, and hearing following restoration of bilateral hearing with a CI.

Goals: Children with SSD using a CI provide a unique model for studying how restored input shapes binaural cortical development. Early auditory deprivation in these children may lead to asymmetric cortical organization and limit binaural integration. The binaural interaction component (BIC) of cortical auditory evoked potentials (CAEPs) is an objective biomarker of binaural cortical processing. The factors underlying its emergence and association with auditory experience and cortical plasticity in children with SSD using a CI, remain unclear. The current study aimed to investigate neurophysiological and clinical factors associated with BIC, focusing on cortical maturation (P1 latency), CI use patterns, and speech perception outcomes.
Material and Methods: Children with pre/perilingual SSD using a CI underwent two CAEP recordings separated by ~4 years. Most prevalent etiologies were cCMV and meningitis. Background and clinical variables included age at CI, duration of CI use, daily CI use (hours), and monosyllabic word and phoneme recognition scores. CAEPs were recorded to a speech token presented via a loudspeaker at 65dBSPL in 3 listening conditions: normal hearing (NH, CI off), CI (with masking to the NH ear) and bilateral (NH+CI). Longitudinal cortical maturation was assessed by change in P1 latency between recordings. Participants were grouped based on presence/absence of BIC.
Results: Children with BIC (58% of the cohort) showed significantly greater daily CI use (p < 0.001) and shorter P1 latencies at the second recording (p < 0.001). A strong negative correlation was found between BIC presence and longitudinal change in P1 latency (r = -0.78, p = 0.004). Duration of CI use was not associated with BIC presence, and no significant differences were observed in age at CI or speech perception outcomes.
Conclusions: BIC presence in children with SSD is associated with increased daily CI use and enhanced cortical maturation, but not with age at CI, duration of CI use or speech perception outcomes. These findings support a bidirectional association between auditory experience and neural plasticity in which greater daily CI use may promote cortical development, while more mature cortical processing may, in turn, facilitate enhanced device acceptance and use. Neurophysiological biomarkers (P1, BIC) serve as sensitive indicators of binaural integration, highlighting their role in guiding clinical follow up and management in children with SSD using a CI.

Goals
To investigate the development of sentence-level semantic processing during the first year after cochlear implantation using the N400 as an objective neural marker.
Material and Methods
Nineteen postlingually deafened adult CI users (mean age: 62 years; contralateral PTA4: 70 dB HL) were tested at 4 weeks and 12 months after initial activation. Sentence comprehension was assessed using an N400 paradigm designed to probe different levels of semantic integration relevant for everyday speech understanding. Sentence endings were (A) semantically correct and contextually expected, (B) semantically correct but contextually unlikely, or (C) semantically incorrect. EEG was recorded using 32 electrodes, and analyses were restricted to correctly classified trials. Data were compared to 19 age-matched normal-hearing controls (mean age: 62 years; PTA4: 17 dB HL).
Results
Normal-hearing controls showed robust N400 effects across all comparisons (A vs. C, B vs. C, A vs. B). CI users exhibited a significant N400 effect for semantically incorrect versus correct expected endings (A vs. C) already at 4 weeks post-activation; however, peak latencies were delayed compared to controls and decreased over time. For correct but unlikely versus incorrect endings (B vs. C), controls again showed a clear effect, whereas in CI users this effect reached significance only after 12 months and remained delayed. For expected versus unlikely correct endings (A vs. B), controls showed a reliable N400 effect, but no significant effect was observed in CI users at either time point.
Conclusion
CAEP measures allow objective tracking of sentence-level semantic processing after cochlear implantation. CI users show early access to coarse semantic processing, as reflected in the differentiation between correct and semantically incorrect sentence endings. In contrast, finer contextual differentiation remains absent or weak, even after one year. This dissociation indicates distinct developmental trajectories of coarse and fine semantic processing. Delayed N400 latencies further suggest less efficient and temporally altered processing compared to normal-hearing listeners.

Goals:
In patients with auditory implants, late auditory evoked potentials (LAEPs) allow the objective investigation of speech processing at the cortical level. Despite their diagnostic potential, LAEPs are rarely used in clinical practice due to methodological and practical challenges. This study aimed to define a measurement paradigm and establish LAEP reference values to enable the assessment of speech processing at multiple levels—from speech-sound detection to semantic processing—within a clinical setting.

Materials and Methods:
Cortical speech processing was investigated in patients with cochlear implants (CI) and normal-hearing (NH) controls, focusing on both speech-sound detection and semantic processing. Short consonants (/m/, /g/, /t/) and sentences were presented, with sentences being either semantically congruent or incongruent. Neural responses were recorded simultaneously using a combined setup of a 72-channel EEG research system (NeurOne, Bittium, Finland) and a two-channel clinical system (Eclipse EP25, Interacoustics, Denmark). LAEPs elicited by consonant and sentence onset, as well as by the sentence-final word, were compared across recording systems, conditions, and groups.

Results:
LAEPs to consonant and sentence onsets, reflected in the N1–P2 complex, and the N400, reflecting semantic processing, were observed. While these responses were present in both recording systems, data from the two-channel clinical system were considerably noisier in CI users, likely due to CI-related electrical artifacts.

Conclusion:
Our findings demonstrate that speech processing in CI patients can be assessed at multiple levels—including speech sound detection and semantic processing—using LAEPs, even with a limited clinical measurement setup. By carefully selecting stimuli, conditions, and electrode positions, meaningful insights into cortical speech processing can be obtained in CI users in routine clinical practice, although strategies to further reduce CI-related artifacts remain necessary.

Goals: Although many cochlear implant (CI) users achieve good speech comprehension, music perception often remains challenging. Perceiving subtle musical structures, such as timbre, and identifying musical instruments can be difficult for CI users [1,2]. This study investigates whether CI adjustment using MED-EL's anatomy-based fitting (ABF) has an effect on the ability to discriminate timbre. The frequency bands of the CI electrodes are individually adjusted based on postoperative imaging data. This could improve musical perception for CI users. This study aims to investigate the influence of ABF on CI users’ ability to discriminate timbre, employing both objective and subjective methods.
Material and Methods: The study focuses on adult CI users with at least six months of CI experience and includes a control group of age-matched individuals with normal hearing. Electroencephalography (EEG) recordings are obtained using a mismatch negativity (MMN) paradigm. The multi-feature paradigm [3] employed in this study is based on the Alberti bass, a well-known musical sequence. The timbre is changed at predefined time points within the sequence. The corresponding MMN amplitude and latency are quantified as measures of pre-attentive auditory discrimination. In addition, psychoacoustic tests are performed to assess timbre discrimination and instrument recognition. The CI users undergo all tests once with ABF settings and once with existing clinical default frequency assignment. The control group without CI performs all measurements once.
Results: Initial pilot measurements and trends regarding the effect of ABF on timbre discrimination will be presented. MMN amplitude and latency will be compared regarding fitting strategy. Behavioral psychoacoustic outcomes will be reported and examined for convergence with the patterns observed in the MMN measures.
Conclusion: This study will provide objective and behavioral data on whether anatomy-based fitting enhances timbre discrimination in CI users. Demonstrating improved auditory discrimination under ABF could inform CI fitting strategies aimed at improving music perception.

Goals:
Artifacts generated by cochlear implants (CIs) can interfere with the interpretation of neurophysiological data by overlapping neural signals in EEG recordings. This study systematically investigated the impact of key technical parameters, such as speech coding strategy, pulse width, and amplitude range, on CI-induced EEG artifacts. Two phantom models, a saline-filled container and a pumpkin model, were employed to determine whether a simple and readily constructed model is sufficient for reliable artifact characterization.
Material and Methods:
Two phantoms were developed: a plastic container filled with 0.9 % saline and a hollowed pumpkin filled with the same solution, each implanted with a Nucleus CI522 device. An Easycap 64-channel EEG cap recorded electrical signals from both phantoms. Stimulation parameters were varied systematically, two coding strategies (ACE, MP3000), pulse widths from 25 to 400 μs, and two amplitude ranges (110 – 150 CL and 150 – 190 CL). Two different stimulus types (sine and chirp) were presented via loudspeaker. All data were acquired with a SynAmps RT amplifier (DC mode, 10 kHz sampling) and analyzed using custom Python routines. Artifact features were quantified from filtered and averaged epochs using Hilbert envelope and spectral density analysis.
Results:
Artifact morphology across both phantoms exhibited a distinctively rising onset, plateau, and decay pattern that closely resembled the envelope of the acoustic stimulus. Artifact amplitude demonstrated a linear correlation with pulse width (slope: 11.59 ± 0.32 V·s⁻¹), while increased stimulus amplitudes resulted in larger artifacts with consistent morphology. The coding strategy primarily influenced the plateau and decay phases, with MP3000 producing lower but noisier artifacts compared to ACE. Power spectra revealed peaks at multiples of the stimulation rate, irrespective of the stimulus type. The type of phantom had a negligible impact on artifact structure, although absolute amplitudes varied.
Conclusion:
The experimental setup and phantom models reliably characterized CI-induced EEG artifacts. Both the basic saline container and the geometrically structured pumpkin phantom proved suitable for artifact investigations, with the former providing a resource-efficient reference for future technical assessments. These findings support the use of simple phantom setups for systematic exploration and prediction of CI-related EEG artifacts.

Goal
For people that use cochlear implants (CIs) to hear, electric current spread and interactions between spectral channels are key factors that limit speech perception. There exists no method yet to mitigate these factors for CI speech perception. Due to the reduced signal resolution in CI hearing, speech perception likely requires high cognitive effort to work out what has been said, especially in noisy situations. However, the relationship between signal resolution and cognitive demand is not well understood for speech-in-noise perception with CIs.
Method
We studied the causal interaction between signal resolution and cognitive effort during speech perception in background noise. First, we used spectral blurring to causally manipulate signal resolution and combined this with (1) a novel N-Back speech recognition task in background noise to manipulate cognitive effort and (2) a manipulation of neural spread of excitation by adapting the presentation level. Two groups performed the tasks using a counter-balanced within-subject design: 12 normal-hearing (NH) adults using vocoder-based CI simulations and 12 adult CI recipients.
Results
There were highly significant effects of spectral blurring on speech-in-noise perception across groups and participants. The NH group performed slightly better than the CI group. The cognitive manipulation also significantly affected speech recognition scores in noise for both groups. However, only for the CI group there was a significant interaction effect between spectral blurring and cognitive effort. The CI group showed a change in susceptibility to higher spectral blurring amounts under increased cognitive effort, which supports the assumption that signal resolution and cognitive demand influence each other in a complex manner with an upper limit of resilience. The level manipulation did not lead to a change in performance and there was no interaction with spectral blurring, suggesting that this manipulation was not effective.
Conclusion
We found that signal resolution interacted with cognitive demand in the CI group but not in the NH group. These findings support the assumption that multiple factors combine and interact to define the limits of speech perception success with CIs, including background noise, signal degradations and cognitive effort. The observed differences between individual listeners and between the two groups indicate varying degrees of resilience to different sources of perceptual difficulty.

Goals: This study evaluated interpeak latency (IPL) and its inter-trial variability (IPLVAR) of the electrically evoked compound action potential (eCAP) as potential alternatives to the phase-locking value (PLV) for quantifying cochlear nerve (CN) synchrony in cochlear implant (CI) users.
Design: The IPL and IPLVAR were assessed in 28 postlingually deafened adults, 37 children with auditory neuropathy spectrum disorder (ANSD), 43 children with cochlear nerve deficiency, and 38 children with typical sensorineural hearing loss. All participant groups include both male and female participants. Their associations with temporal resolution and speech perception outcomes were evaluated in children with ANSD and/or adult CI users. Frequency analysis of eCAP recording noise was conducted to understand its impacts on synchrony metrics. Simulations of inter-trial jitter in the eCAP were performed to quantify how the synchrony metrics varied with increased temporal jitter, recording noise level, and sampling rate.
Results: eCAP traces recorded in all patient groups showed a multiple-peak point issue affecting the accuracy of IPL and IPLVAR assessments. Temporal resolution and speech perception outcomes were significantly correlated with IPLVAR but not with IPL metrics. Simulation results showed that the IPL was less sensitive to inter-trial temporal jitter than the IPLVAR and the PLV. The reliability of the IPL improved slightly at higher sampling rates. In contrast, the PLV was reliable across a range of sampling rates, with just moderate changes in the shape of the PLV versus jitter relationship, and the IPLVAR metric was largely unaffected by the sampling rate. Increasing recording noise decreased the mean PLV values and increased the mean IPLVAR values without affecting the reliability of these metrics, except at very low SNRs for the IPLVAR. In contrast, the reliability of the IPL degraded substantially with increasing noise. Increasing the sampling rate reduced the effects of recording noise on all three metrics, although only to a limited extent.
Conclusions: The IPL is not a reliable indicator of CN synchrony. The IPLVAR reflects neural synchrony in the CN but is more susceptible to high levels of eCAP recording noise than the PLV. Increasing sampling rate enhances the PLV sensitivity, whereas the IPLVAR is reliable across a range of sampling rates. Overall, the PLV remains the preferred metric for quantifying neural synchrony in CI users.

Goals: This study aimed to develop a clinical tool using onset cortical auditory evoked potentials (CAEPs) elicited by pure-tone bursts to estimate ear- and frequency-specific audiometric thresholds in children with auditory neuropathy spectrum disorder (ANSD) and children with typical sensorineural hearing loss (SNHL).
Design: In this newly developed testing paradigm, tone-burst stimuli with octave frequencies between 250 and 4000 Hz were presented at −20, −10, 0, 5, 10, 20, and 40 dB sensation level (SL; re: behavioral threshold), or at 120 dB SPL (maximum output level) when the desired sensation level could not be reached due to severe hearing loss. Stimuli were presented using a continuous presentation paradigm in which tone bursts with different frequencies and stimulation levels were delivered in random order with an interstimulus interval randomly jittered between 400 and 800 ms. Eye-blink artifacts in the electroencephalogram (EEG) were removed using the iterative template-matching method described by Valderrama et al. (2018). The peak phase-locking value (PPLV) was calculated for each frequency and stimulation level between 1–45 Hz within a time window in which onset CAEPs are expected in children with ANSD (20–250 ms after stimulus onset) using the method reported by Mao et al. (2018). A CAEP response was considered present if the PPLV distributions for the test level and baseline level (−20 dB SL) overlapped by less than 5% (Mao et al., 2018). The CAEP threshold was defined as the lowest stimulation level at which a CAEP was detected. To date, CAEPs recorded using the continuous presentation paradigm from four children with SNHL and eight children with normal hearing, as well as CAEPs recorded using the conventional discrete presentation paradigm from five children with ANSD, have been analyzed. For each participant, CAEP thresholds were compared with behavioral thresholds obtained using pure-tone audiometry.
Results: CAEP thresholds were highly correlated with behavioral thresholds across all three participant groups, particularly at higher frequencies. Differences between objective CAEP thresholds and behavioral thresholds varied across participants and octave frequencies.
Conclusions: Preliminary results suggest that this newly developed clinical tool can be used to estimate ear- and frequency-specific audiometric thresholds in children, including those with ANSD.

Goals: This study aimed (1) to determine the biological basis of the interphase gap (IPG) effect and (2) to identify the most informative parameters, analytical methods, and quantitative scales for evaluating the IPG effect in human cochlear implant (CI) users.
Design: The IPG effect was quantified using multiple parameters, analytical methods, and quantitative scales (three combinations using linear or logarithmic scales for the input and output variables) across three pediatric CI groups with differing cochlear nerve (CN) anatomy: children with cochlear nerve deficiency (CND), GJB2 mutations, and idiopathic sensorineural hearing loss (SNHL). All participants in the GJB2 and idiopathic SNHL groups had normal-sized CNs (NSCNs) in the test ear. Neural synchrony, a property depending on neural health, was assessed using phase locking value (PLV) and compared between children with CND and those with GJB2 mutations.
Results: The PLV did not differ significantly between the CND and GJB2 groups, nor did it correlate with the IPG effect in GJB2 cases, regardless of parameter, analytical method, or quantitative scale. In contrast, consistent group differences in IPG effects on stimulation level offset and maximum slope of the eCAP input/output function were observed across all analytical methods and quantitative scales. The sensitivity of other eCAP measures—threshold, maximum amplitude, and overall linear slope—to group differences varied by quantitative scale.
Conclusions: The IPG effect primarily reflects the number of active CN fibers rather than their health. Stimulation level offset and the IPG effect on maximum slope are robust indicators of CN fiber counts in CI users and are unaffected by the choice of quantitative scale.

Background:
Objective electrophysiological measures, including electrically evoked auditory brainstem responses (eABR), play an important role in determining cochlear implant (CI) candidacy in children with auditory neuropathy spectrum disorder (ANSD) and suspected cochlear nerve deficiencies

Methods:
Two paediatric ANSD cases underwent assessment including newborn ABR, behavioural audiology, CT/MRI (where possible), hearing‑aid evaluation, MDT review, and diagnostic promontory stimulation eABR.

Results:
Case 1: MRI demonstrated bilaterally hypoplastic cochlear nerves, CT showed absent right cochlear nerve canal and a possible reduced left cochlear nerve. Hearing aids provided no functional responses. Patient was referred for Promontory eABR with the support from MEDEL UK. The left eABR showed Wave V (Type C) response. The right eABR showed no response up to maximal levels. Outcome: Listed for left CI with Otoplan.

Case 2: MRI contraindicated due to pacemaker. CT indicated bilateral cochlear malformations and uncertain nerve integrity. Hearing aids provided no functional response. Patient was referred for Promontory eABR with the support from MEDEL UK. The left eABR showed no response up to maximal levels. The right eABR also showed no response up to maximal levels with a myogenic potential seen. Outcome: Referred for ABI assessment.

Conclusion:
Promontory eABR helped to clarify cochlear nerve status where imaging and behavioural results were either unobtainable or inconclusive. Objective electrophysiology supported CI vs ABI differentiation, reducing delay in onward referral and family uncertainty. Additionally, this prevented a long period of uncertainty with Case 2 in terms of cochlear implantation rehabilitation and trying to differentiate between the sometime challenging question of cochlear implant benefit and developmental milestones in early years.

Goals: Despite decades of research, objective fitting methods for cochlear implants (CI) have yet to be established in routine clinical practice. Although the stapedius reflex (SR) threshold has been shown to be particularly well suited as a postoperative upper limit for the maximum comfort level (MCL), the integration of an appropriate technique into the implant system remains technically challenging. Especially infants and other users with limited communication skills could be benefit from objective determination of the MCL. This study demonstrates that monitoring middle ear pressure can identify acoustically and electrically evoked stapedius reflexes in patients and CI implanted sheep.
Methods: In eleven patients, a solid-state pressure transducer (Micro-Cath, Millar, USA) was placed in the middle ear during a surgical tympanotomy. Contralateral an acoustic stimulation was delivered using a tympanometer (eTymp, Biomed Jena, Germany), and the resulting pressure changes were recorded. Additionally, nine Merino sheep underwent surgical mastectomy and acute CI electrode insertion (MED-EL, Austria). After implantation, the pressure sensor was placed in the middle ear and electrical stimulation bursts of increasing intensity were delivered through CI contacts while middle ear pressure was recorded. In addition, one animal was equipped for a chronic experiment with a pressure sensor for half a year and stimulated acoustically at different time points.
Results: We were able to successfully determine the SR in patients and sheep after acoustically respectively electrically stimulation. SR related pressure changes were detected in 10 of 11 patients and 9 of 10 sheep, correlated with tympanometry assessment. Absence of SR in one sheep was confirmed by both methods, indicating true reflex absence rather than detection failure. Heartbeat-induced artefacts were observed in patients and sheep, but the SR signal for sheep was much larger. Moreover, the middle ear sealing in sheep was not as critical as observed in human, but this might be related to the fact that the middle ear was constantly flooded with liquid.
Conclusion: This works contributes to the development of middle ear pressure as an objective measure for SR detection. It shows that it can be implemented with acoustic and electrical stimulation with similar results. The data provides evidence that the method have comparable results with the current gold standard of acoustic stimulation via tympanometer de

Objective Measures of Neural Health in Cochlear Implant Users and Their Correlation with Speech Outcomes

Background:
Neural health plays a pivotal role in determining outcomes in cochlear implant (CI) users. Objective measures such as electrically evoked compound action potentials (eCAP), electrically evoked auditory brainstem responses (eABR) and impedance telemetry which gives information of the integrity of the auditory pathway
Objective
To evaluate the relationship between objective electrophysiology and speech perception outcomes in CI recipients.

Methodology:
A cross-sectional study was conducted on CI users with at least 2 years of device experience. eCAP thresholds and amplitudes were recorded across multiple electrodes to assess peripheral neural responsiveness. eABR testing was performed to evaluate neural synchrony along the auditory brainstem pathway. Impedance telemetry values were obtained to examine electrode integrity and electrode–tissue interface status. Speech perception was assessed in quiet and noise. Correlation analyses were performed between objective measures and speech scores.

Results:
Stronger eCAP responses (lower thresholds, higher amplitudes) and well-defined eABR waveforms were associated with better speech perception scores. Stable and lower impedance values indicated a healthier electrode interface and were linked to improved outcomes. Conversely, absent or elevated thresholds, abnormal eABR morphology, high impedance, were associated with poorer outcomes.

Conclusion:
Objective measures of neural health demonstrate significant correlations with speech perception in CI users. A combined assessment using eCAP, eABR and impedance telemetry provides a comprehensive understanding of auditory function. Incorporating these measures into routine clinical practice can enhance individualized programming and improve prognostic accuracy.

Goals
Defining a set of stimuli that can be distinctly characterized by their frequency content is a first step towards developing an objective method of assessing a person’s hearing capacity using the fre-quency following response (FFR). Phase-locking limitations of the auditory nerve impose limits on the types of sound stimuli that can be meaningfully applied. This report explores some vowel sounds and their representation in the FFR.
Material and Methods
FFR recordings involving vowel stimuli with formants below 2kHz were measured (with a DUET EEG system kindly loaned from Zeisberg GmbH/Intelligent Hearing Systems) with 3 normal hear-ing volunteers. Live-voice recordings of 100ms Ling vowels /a/ /i/ /u/, plus 100ms synthetic vowels /a/ /i/ /ɔ/ and /u/ were tested. No-stimulus recordings were also made to obtain the noise profile of the measurement amplifier. All recordings were then bandpass filtered (70-1500Hz) to reduce low frequency and high frequency noise. Additional noise reduction techniques such as spectral pre-emphasis, spectral subtraction and harmonic comb filtering were investigated to examine their effects on the stimulus F1 & F2 formant estimates within the FFR recordings using Praat software.
Results
F0 is very well and accurately represented in the FFR recordings. Higher F0 harmonic content present within both stimulus and recordings, together with measurement noise, generally made it more difficult to extract formant information from the FFR recordings. This appeared to be more so for live voice recordings compared to synthetic. Neither harmonic comb filtering at F0 of the input signal nor spectral subtraction of the amplifier pink noise performed offline helped to reduce the noise substantially. Pre-emphasis performed the poorest of the noise reduction techniques at-tempted. The noise reduction techniques were therefore not very effective in helping to improve formant estimates from the FFR recordings. Not all stimuli had their formants well represented within the FFR recordings.
Conclusion
Vowel stimuli characterized by their formant contents in FFR recordings should have maximum separation along at least one dimension like /a/ and /u/ for F1, with synthetic vowels preferred over live voice vowel recordings. Post-processing to reduce noise within the FFR recordings have mini-mal effect. These together form the basis of further investigations using the FFR to objectively characterize the auditory capacity of individual listeners.

Introduction
Inner ear malformations can be challenging cases for electrode array placement. In case of common cavity malformation, placement can be especially challenging due to the undifferentiated cavity formation. This study describes a case who received sequential bilateral cochlear implants (CIs). The procedure used in this case shows nicely how intra-operative imaging, such as fluoroscopy during insertion of the electrode array and objective measures, here, EABRs, can be used to optimise electrode array placement and finding areas with best responses.
Method
Pre-op CT scans were analysed to measure the size of the cochlea and common cavity width. Different test electrodes were used to find the best fit for this cochlea. Before insertion Round Window Electrically Evoked Auditory Brainstem Response (RW-EABR) measurement was carried out to investigate to investigate if this cochlea will transfer electrical stimulation along the auditory pathways. Using intra-op fluoroscopy, different test electrodes were inserted to find the best fit for this cochlea. After insertion of chosen electrode array, EABR measurements were done to measure the areas inside the cochlea with best response. In the post-operative phase, a suitable frequency allocation table was defined based on intra-cochlea placement. After a successful first surgery, a second sequential surgery was carried out, with placement sited as symmetrically as possible. For the second CI, the same measurements and imaging tasks were carried out as with the first implant. Frequency allocation tables were matched to both sides to avoid frequency mismatch.
Results
After the first surgery, the patient responded well to environmental sounds and therefore a second surgery was carried out. There were concerns that a second CI may cause a mismatch of sounds causing a negative effect in sound perception. Due to a similar placement as possible and adapting the frequency allocation table the patient could benefit from both CIs.
Conclusion
This case study shows a surgical procedure of a common cavity that optimizes the placement of electrode arrays and adapting frequency allocation tables. Additionally, EABR measurements indicate areas inside the common cavity with better responses.

Objective. Is wider pulse duration an option to reduce the spread of excitation?
Methods. This is a prospective cross-sectional study approved by the Ethical Board of the Institution under protocol number 188.167.
Intraoperative recordings of the neural response telemetry thresholds (tNRT) and spread of excitation (SOE) functions of adults and children who received straight array cochlear implants (CI 422, CI 522 or CI 622) were collected in electrode contact 11. tNRTs were assessed with AutoNRT algorithm, with default pulse width (25s), and with 37s in advanced mode. SOE functions were recorded at least 10 current units above tNRT, with 25s pulse width and 37s a function of eleven different masker electrodes. Custom Sound® EP software provides SOE width in millimeters, and for this study the width was collected at 50%, 75% and 90% transection levels. The tNRT, stimulus level to assess the SOE function, amplitude of the stimulation electrode and the peak of the function, as well as the SOE width were compared by the paired t-test with 5% of significance level.
Results. Thirty-five recordings, 20 from children and 15 from adults were collected.
There was a significant difference of the tNRT and consequently the simulation level used to assess the SOE function, between stimulus of 25 µs and 37 µs pulse widths, in adults and children.
There was no difference on the amplitude of the function or the SOE width at any of the transection levels in children or adults.
There was no significant correlation in adults between the stimulus level and the SOE width, nor between the amplitude of the peak electrode and the SOE width for both pulse widths.
In children, there was no significant correlation between the stimulus level and the SOE width, nor between the amplitude of the peak electrode and the SOE width for 37 µs pulse width. However, for 25 µs pulse width, a strong positive correlation was found between the amplitude of the stim electrode and SOE width at 75% (rs = 0.7564, p = 0.0001) and 90% (rs = 0.7454; p = 0.0002). No significant correlation between the stimulus level and the SOE width was found.
Conclusion. Widening the pulse width to 37 µs has a significant impact on the tNRT and consequently on the simulation level used to assess the SOE function, nevertheless there was no influence on the SOE width at the mid portion of the cochlea (electrode 11) at any of the transection levels in children or adults.

Goals
To critically evaluate the clinical utility of routinely used objective measures in cochlear implant (CI) programming and to examine whether they reliably predict auditory outcomes in everyday clinical practice.
Material and Methods
A focused evidence-based synthesis was conducted on routinely used objective measures in CI programming, including electrically evoked compound action potentials (eCAP), impedance telemetry, and electrically evoked stapedius reflex thresholds (ESRT). Relevant peer-reviewed studies and clinical reports were analysed to assess their role in guiding programming and their relationship with behavioural and speech perception outcomes.
Results
Routine objective measures provide essential physiological insights into neural responsiveness, electrode integrity, and stimulation levels, particularly in populations where behavioural responses are limited. However, across studies, their ability to predict functional auditory outcomes remains inconsistent. While eCAP and ESRT can support initial programming and approximate stimulation parameters, their correlation with speech perception outcomes is variable and often weak. Impedance telemetry is indispensable for device monitoring but does not directly inform performance outcomes.
Conclusion
Routine objective measures are integral to cochlear implant programming, yet their role as predictors of clinical outcomes is limited. Rather than definitive indicators, they function as supportive tools within a broader, behaviourally driven framework. This underscores the need for cautious interpretation and reinforces that optimal CI outcomes depend on integrating objective data with patient-specific behavioural measures.