gms | German Medical Science

The observed interindividual variability in CI users" speech perception has not been explained accurately so far. One potential factor is the number of surviving spiral ganglion neurons (SGN) and the degree of their degeneration, known as neural health. Studies show exploratory results in terms of correlation between neural health measures and speech intelligibility [1]. Therefore, this study aimed at investigating this relationship in more detail by taking into account the speech band importance function [2] and demographic data.
A group of 13 bilaterally implanted subjects (26 ears) were included in the study. eCAP recordings were made in both ears using anodic leading and cathodic leading biphasic pulses in monopolar mode. Both forward masking and alternating polarity approaches were used to reduce the artifact. In order to estimate the neural health in the region of each stimulating electrode, the change in the slope of the amplitude growth function as a result of increasing the interphase gap (IPG) from 2.1 µs to 30 µs was measured and defined as the IPG effect (IPGE) on slope [3]. Higher IPGE on slope is associated with better neural health. To compare this regional measure with monotic speech perception, the relative importance of each frequency band for speech intelligibility
in noise in different cochlear regions was also taken into account. This was implemented by applying a weighting function [2] to the IPGE on slope measured at every electrode contact in the cochlea. Standard and weighted Pearson correlation analysis was applied to investigate the relationship between IPGE on slope and speech intelligibility as well as between IPGE on slope and demographic data of type age at implantation, duration of hearing loss, hearing aid experience and CI experience.

The results showed a small but significant correlation (R2 = 0.24, p< 0.05) between IPGE on slope and speech intelligibility for the weighted Pearson correlation. This correlation was increased (R2 = 0.32, p< 0.01) after applying the speech band importance weighting function. A larger, significant and negative correlation was also found between IPGE on slope and age at implantation (R2 = 0.51, p< 0.001) whereas the negative correlation between IPGE on slope and hearing aid experience was low (R2 = 0.17, p< 0.05). No correlation was observed between IPGE on slope and either duration of hearing loss or CI experience.

Based on the above-mentioned findings it can be concluded that IPGE on slope shows the potential for a clinically relevant determination of neural health and consequently for partially explaining the interindividual variations in speech intelligibility.