12Intraoperative CI

The growth area of clinical ECochG, and the only one with randomised trial evidence. Real-time monitoring of the cochlear microphonic via the implant's own back-telemetry detects insertion trauma as it happens, and triggering surgical intervention on a >=30% CM amplitude drop preserves residual hearing — a level 1b finding from Campbell et al.'s 2022 single-blinded RCT.

Hearing preservation during cochlear implantation matters because patients with residual low-frequency hearing benefit substantially from electroacoustic stimulation — the CI processes high frequencies electrically while natural acoustic hearing carries the lows, giving better speech-in-noise and music perception than electrical-only stimulation. Loss of residual hearing during the insertion negates this benefit. Insertion trauma is the principal cause, and until very recently the surgeon had no real-time information about whether the array was injuring the cochlea or not.[2021]Intraoperative ECochG closes that feedback loop, and Campbell et al.'s 2022 RCT is the level 1b anchor for the field.[2022]

FWhy monitor at all

The case for intraoperative monitoring rests on three observations:

1. Insertion trauma is common. Even with experienced surgeons and soft-tip electrodes, a substantial minority of insertions damage the basilar membrane or scala vestibuli/tympani transition.[2021, 2025]
2. The trauma happens in seconds. Acute CM amplitude drops on the order of seconds occur during insertion and predict more severe postoperative hearing losses.[2022, 2012, 2014]
3. Intervention can reverse the trauma. Pausing the insertion, withdrawing the array a few millimetres, repositioning, and re-advancing can restore the CM amplitude. If the trauma is mechanical and acute (not yet causing permanent damage), the cochlea recovers.[2022]

TRecording via the implant itself

The key enabling technology was Campbell et al.'s 2015 demonstration that the cochlear implant's own back-telemetry amplifier can record ECochG from its own electrodes during the implantation.[2015] Before this, intraoperative ECochG required a separate transtympanic or round-window electrode — feasible but cumbersome.[2012, 2014] Recording intracochlearly via the CI means:

• Substantially larger amplitudes than extracochlear recording. The intracochlear electrode is close to the generators, giving CM in the tens to hundreds of microvolts rather than the microvolt-scale tiptrode signals.[2021]
• Real-time during insertion — no extra hardware in the surgical field, no time penalty.
• A moving recording site. As the array advances, the recording electrode tip moves apically along the cochlea, which changes which generators contribute to the recorded potential. Algorithms must account for this.[2021]

The recording protocol uses a low-frequency acoustic stimulus (250 or 500 Hz tone burst) delivered through the patient's ear canal during surgery. The CM is extracted from the recording, its amplitude tracked over time, and a drop > 30% from the prior maximum triggers an alarm.[2022, 2022] ECochG-TR via the implant is also useful in characterising residual cochlear function in auditory neuropathy candidates before insertion proceeds.[2017]

TInsertion animation

The schematic below shows what intraoperative ECochG looks like in practice: a CI electrode array partially inserted into the cochlear spiral, with three CM amplitude trends drawn alongside corresponding to the canonical Growth, Fluctuating, and Total Loss response patterns. The Campbell 2022 RCT defines a CM amplitude drop of ≥30% from the prior running maximum as the trigger for surgical intervention.[2022]

Pick a pattern, then drag the depth slider or press Play to advance the electrode along the cochlear spiral. The CM amplitude trace tracks insertion depth on the right; the running maximum is shown as a faint dashed reference, and the trace recolours and an alarm panel appears the moment the current amplitude drops ≥ 30% from the running peak — the Campbell 2022 RCT intervention trigger. Total Loss reaches that threshold around 60% depth; Growth never does.[2022]

TThree response patterns

Haumann and colleagues (2022) identified three characteristic ECochG response patterns during electrode insertion in 47 adult patients with measurable residual hearing.[2022] Each has different prognostic implications.

CThe Campbell 2022 RCT

The clinical-evidence centrepiece of this module is Campbell et al.'s 2022 randomised controlled trial in Hearing Research.[2022] Sixty adults receiving the Cochlear Thin Straight Electrode were randomised to:

The intervention group showed significantly better hearing preservation than the control group at the primary endpoint, with the protocolised intervention reversing acute CM drops during the procedure. This is the first level 1b evidence that intraoperative ECochG-triggered intervention preserves residual hearing — and the foundation for the technique's emerging place in mainstream CI surgery. Trecca et al.'s recent systematic review summarises the broader literature and positions the Campbell trial as the field's evidentiary anchor.[2025]

CBeyond amplitude — the modern direction

The most recent multicentre work (Andonie et al. 2025, Melbourne–Bern–Zurich) extends the simple amplitude-drop trigger into a richer multimodal feature set:[2025]

• CM amplitude events— the original ≥30% drop trigger, kept as the primary feature.
• ANN:CM amplitude ratio — the auditory nerve neurophonic (ANN, the neural component) relative to the CM. Changes in this ratio reflect changes in neural contribution that pure CM amplitude misses.
• CM phase tracking — phase shifts can indicate destructive interference vs genuine trauma, helping resolve the ambiguity in the Fluctuating pattern.
• Event evaluation paradigms — four ways of weighting CM events: naive (all events count), deep (only after basal electrode is in), persistent (only non-recovering drops count), and combined (deep + persistent).

Multifrequency monitoring is also moving from research into clinical use. Walia et al. (2022) showed that 250 Hz CM drops are more predictive of hearing preservation than 500 Hz; combining both gives the best feedback. When multifrequency is not available, the default has shifted from 500 Hz to 250 Hz on the basis of this work.[2022]

FClinical case

TCSelf-assessment