# Electrocochleography for Monitoring Hearing Preservation During Cochlear Implantation

## TL;DR

This study shows that automated electrocochleography during cochlear implant surgery can predict hearing preservation, helping surgeons avoid damaging residual hearing.

## Contribution

The study introduces automated analysis of intraoperative electrocochleography as a feasible method for real-time hearing preservation monitoring during cochlear implantation.

## Key findings

- Persistent or late cochlear microphonic amplitude decreases during implant insertion were strongly linked to postoperative hearing loss.
- A rising amplitude ratio between auditory nerve neurophonic and cochlear microphonic signals was associated with better hearing preservation.
- Stable cochlear microphonic phase also showed a trend toward improved hearing outcomes.

## Abstract

This cross-sectional study assesses whether automated analysis of intraoperative cochlear microphonic amplitude decreases is associated with hearing preservation.

Is automated intraoperative monitoring of residual hearing during cochlear implantation feasible using objective electrocochleography analysis?

In this cross-sectional study of 112 adults receiving cochlear implants, decreases in cochlear microphonic amplitude were associated with postoperative hearing loss when they were persistent or occurred near the end of electrode insertion. Hearing preservation was particularly poor when the amplitude ratio between the auditory nerve neurophonic and cochlear microphonic did not increase or when the cochlear microphonic phase remained stable.

Automatically extracted electrocochleography features may predict residual hearing outcomes during cochlear implantation, and these findings support the development of real-time feedback systems assisting surgeons during implantation.

Many cochlear implant candidates retain residual hearing, which can improve speech perception and overall outcomes. However, residual hearing is frequently lost during implantation. Intraoperative electrocochleography may enable monitoring of residual hearing, but its clinical relevance is currently limited due to reliance on expert interpretation, the absence of real-time analysis algorithms, and signal variability.

To assess whether automated analysis of intraoperative cochlear microphonic amplitude decreases (events) is associated with hearing preservation and to evaluate additional electrocochleography features enhancing potential predictive performance.

This multicenter, cross-sectional cohort study included adults receiving a cochlear implant with a preoperative hearing threshold not greater than 85 dB hearing level at 0.5 kHz and cochlear microphonic amplitude of 5 μV or higher across 3 tertiary referral centers in Melbourne, Australia; Bern, Switzerland; and Zurich, Switzerland. Intraoperative real-time electrocochleography and impedance data were collected between 2017 and 2025, with audiometric follow-up 3 months postoperatively.

Cochlear implantation with the Slim Straight Electrode array (Cochlear Limited).

Primary outcome was binary hearing preservation at 3 months postoperatively, defined as less than 25% loss relative to the preoperative low-frequency pure-tone average at 0.25 to 1 kHz. Primary variables assessed as possible predictors included magnitude, location, and persistence of cochlear microphonic events. Post hoc analysis explored associated changes in the auditory nerve neurophonic:cochlear microphonic amplitude ratio and cochlear microphonic phase.

A total of 112 patients met inclusion criteria and were analyzed (median [IQR] age at surgery, 68 [58-75] years; 57 [51%] female and 55 [49%] female). Cochlear microphonic events persisting or occurring near the end of insertion were associated with loss of residual hearing (adjusted odds ratio, 31.58 [95% CI, 6.36-205.36] and 52.96 [95% CI, 8.02-472.63], respectively), independent of age, preoperative hearing, and participating hospital. Events with rising amplitude ratio between auditory nerve neurophonic and cochlear microphonic were associated with better hearing preservation (mean difference in hearing preservation, 24.4% [95% CI, 7.3%-41.5%]). Stable cochlear microphonic phase showed a similar trend (mean difference in hearing preservation, 20.9% [95% CI, 2.9%-38.9%]).

This cross-sectional study demonstrates the feasibility of automated intraoperative electrocochleography for possible prediction of hearing preservation during cochlear implantation. Persistent cochlear microphonic events near the end of insertion were associated with hearing loss. Additional electrocochleography features may improve signal interpretation. These findings support the development of real-time feedback systems to guide cochlear implantation.