Towards personalized auditory models: predicting individual sensorineural-hearing-loss profiles from recorded human auditory physiology

Abstract

Over the past decades, different types of auditory models have been developed to study the functioning of normal and impaired auditory processing. Several models can simulate frequency-dependent sensorineural hearing loss (SNHL), and can in this way be used to develop personalized audio-signal processing for hearing aids. However, to determine individualized SNHL profiles, we rely on indirect and non-invasive markers of cochlear and auditory-nerve (AN) damage. Our progressive knowledge of the functional aspects of different SNHL subtypes stresses the importance of incorporating them into the simulated SNHL profile, but has at the same time complicated the task of accomplishing this on the basis of non-invasive markers. In particular, different auditory evoked potential (AEP) types can show a different sensitivity to outer-hair-cell (OHC), inner-hair-cell (IHC) or AN damage, but it is not clear which AEP-derived metric is best suited to develop personalized auditory models. This study investigates how simulated and recorded AEPs can be used to derive individual AN- or OHC-damage patterns and personalize auditory processing models. First, we individualized the cochlear-model parameters using common methods of frequency-specific OHC-damage quantification, after which we simulated AEPs for different degrees of AN-damage. Using a classification technique, we determined the recorded AEP metric that best predicted the simulated individualized CS profiles. We cross-validated our method using the dataset at hand, but also applied the trained classifier to recorded AEPs from a new cohort to illustrate the generalisability of the method.

The variability in potential biomarkers for cochlear synaptopathy after recreational noise exposure

“…A study by the same lab could not find noise-exposure as a predictor of EFR magnitude (Prendergast et al, 2019). In contrast, animal studies (Möhrle et al, 2016; Parthasarathy and Kujawa, 2018; Shaheen et al, 2015) and those that used computational modelling of the auditory periphery or human data have shown the possible value of EFRs in diagnosing CS (Bharadwaj et al, 2015; Keshishzadeh et al, 2020; Keshishzadeh et al, 2021; Paul et al, 2017; Vasilkov et al, 2021).…”

“…Twenty millisecond epochs were extracted relative to the stimulus onset and baseline correction was applied by subtracting the mean-value of each epoch. Two hundred epochs, equal number of each polarity, with the highest peak-to-trough values were rejected and the remaining 4800 epochs were averaged (Keshishzadeh, Garrett, & Verhulst, 2020). Mean ABR level series, starting from 100 dB-peSPL and down, were plotted to visually identify the peaks based on the mean latency intervals by Table 8-1 in Picton (2011).…”

“…Exemplar SAM and RAM EFR spectra and corresponding noise-floors are shown in Figure 2A and C, respectively. For detailed explanation of the bootstrapping procedure see Keshishzadeh, Garrett and Verhulst (2020). EFR strength was defined as the summation of the signal-to-noise spectral magnitude at fundamental frequency and the following three harmonics, i.e.…”