13Listening as You Insert: Real-Time Electrocochleography

FThe idea: turning the cochlea into a sensor

A short tone (often a low-frequency click or 250-500 Hz burst) is played into the ear during surgery, and the cochlea's electrical response is recorded from a round-window electrode or, increasingly, directly through the most apical contact of the implant itself. Because the same hardware that will stimulate can also record, the implant becomes a real-time sensor of inner-ear health as it advances - no extra device in the ear canal is needed once intracochlear recording is used. The principle: hair cells and neurons that are alive and undisturbed generate a robust response; trauma to the basilar membrane, lateral wall or organ of Corti shows up as a sudden loss of that response. This builds on the same potentials covered in depth in the Objective Measures chapter - here the focus is their intraoperative, moment-to-moment use rather than diagnostic ECochG.[2017][2009]

TReading the waveform: CM, SP, ANN and CAP

Cochlear microphonic (CM): an alternating-current signal that follows the stimulus waveform, generated mainly by outer hair cells; it is the workhorse of intraoperative monitoring because it is large, fast to acquire and tracks hair-cell health near the electrode tip. Summating potential (SP): a direct-current shift reflecting the nonlinear, mostly hair-cell, response to sustained sound. Auditory nerve neurophonic (ANN): a following response at twice the stimulus frequency arising from phase-locked nerve fibres; the compound action potential (CAP/N1) is the synchronous onset firing of the nerve. In practice the surgeon watches a running amplitude (and sometimes phase) of the CM/ongoing response; the total intracochlear response is larger and cleaner than an extracochlear recording because the contact sits close to the apical, low-frequency region most at risk.[2017][2016]

CThe drop: what an amplitude fall means and what the surgeon does

An abrupt fall in CM amplitude during advancement is the alarm signal - it suggests the array is loading the basilar membrane, has touched the lateral wall, may be tip-folding, or is heading out of scala tympani. Typical reported thresholds for concern are a drop of roughly 30% or more, particularly if it is sustained or occurs late in the insertion; a >=30% reduction predicted poorer 3- and 12-month preservation in a large observational series. Suggested manoeuvres: pause and let the signal recover, withdraw the array a millimetre or two, slow the rate, adjust the trajectory/angle, or stop at a shallower depth - sometimes the CM recovers and insertion can resume. In one real-time telemetry series, 47% of insertions showed a transient or permanent CM drop, and patients whose CM was preserved at the end of insertion averaged about 15 dB better low-frequency preservation.[2016][2020]

CDoes it work - and where it still falls short

The strongest evidence is a randomised trial in which acting on an ECochG drop (pausing or adjusting) gave better residual-hearing preservation than insertion without intervention - moving ECochG from prognostic to genuinely guiding. Limits: a drop does not always equal permanent loss (signals can recover), and a stable signal does not guarantee a perfect outcome - delayed loss can still occur over months from inflammation or fibrosis. Interpretation is confounded by electrode-to-hair-cell distance changing as the array advances, by characteristic-frequency effects, and by movement and electrical artefact; standardised recording protocols are improving reliability. ECochG is one input, not a verdict: it is combined with surgical feel, insertion speed, fluoroscopy/imaging where used, and the recipient's residual hearing to decide how deep and how fast to go.[2022][2020]