9.1 The electrode array and current spread

A cochlear-implant electrode array is a thin (0.4–0.8 mm diameter at the tip, tapering up the shaft), flexible silicone carrier embedded with 16 to 22 platinum-iridium electrode contacts spaced along its length. The array is threaded through a small opening in the round-window membrane or a custom-drilled cochleostomy and pushed into scala tympani — the fluid-filled chamber that runs the length of the cochlear spiral. A well-placed array follows the cochlea’s 2.5-turn helical path for about 16–25 mm, depending on individual anatomy and array length. The most apical (low-frequency) electrode sits roughly at the cochlear position corresponding to ~500 Hz natural hearing; the most basal (high-frequency) electrode sits at the position corresponding to ~8 kHz.

The array’s job is to deliver brief biphasic current pulses to the auditory nerve fibres distributed along the cochlear spiral. The fibres themselves are inside the modiolus — the bony central column around which the cochlea spirals — and their peripheral processes reach radially outward toward the organ of Corti. The current must therefore travel from the intracochlear electrode through the perilymph of scala tympani, through the bony wall of the modiolus, to reach the fibres. The physics of this current path is what sets the channel-resolution limit of the implant.

Current spread in scala tympani

The perilymph filling scala tympani is a roughly homogeneous saline-like medium with bulk resistivity around 0.7 Ω·m. The bony walls of the cochlear duct have a much higher resistivity (typically 100 Ω·m for compact bone). For an electrode delivering current $I$ at a point in scala tympani, the current density at radial distance $r$ in this approximately conductive medium falls off as

$J(r) \;\sim\; \frac{I}{4 \pi r^2}$

for a point-source approximation, or somewhere between $1/r$ and $1/r^2$ for the partially confined geometry of the actual cochlear duct.

The auditory nerve fibres that we want to stimulate sit at varying distances from the array — typically 0.5 to 2 mm from the nearest electrode, depending on the depth of insertion and the individual cochlear anatomy. The current density at the fibre population determines whether the fibre’s threshold is exceeded and the fibre fires. A current pulse of about 50 µA at threshold (clinical “T” level) produces enough current density at the nearest fibres to trigger spikes; a current pulse of about 300 µA at “C” (comfort) level recruits substantially more fibres farther from the electrode.

The clinical implication: the spatial selectivity of an implant electrode is poor. A single electrode at moderate current excites fibres over a 2–5 mm length of the cochlea, blurring what would otherwise be sharp place coding. The 22 electrodes in a typical 22-mm-long array (Nucleus from Cochlear Ltd) cannot deliver 22 independent channels of frequency information — they deliver something closer to 5–8 effectively independent channels because of the overlapping current fields. This is the fundamental limit on CI fidelity.

Monopolar, bipolar, tripolar

Three stimulation modes are clinically available:

• Monopolar (MP). A single intracochlear electrode is the active one; the return electrode is far away, either an extracochlear contact behind the ear or the implant casing itself. The current path is from the active electrode, out through the surrounding tissue, back to the distant return. Current spread is broad — typically 2–4 mm along the cochlear spiral.
• Bipolar (BP). Two adjacent intracochlear electrodes are activated with opposite polarity. The current path is mostly between the two adjacent electrodes, with a more localised current field. Current spread is reduced to ~1–2 mm.
• Tripolar (TP). A central electrode is activated with two flanking electrodes serving as return. The current field is highly localised between the central and the two returns. Current spread is further reduced to ~0.5–1 mm.

Tripolar mode achieves the sharpest current focus and (in laboratory studies) the best channel separation. But sharper focus means lower current density per microamp, requiring higher applied currents to reach threshold — which dramatically increases power consumption and battery drain.

Modern devices ship with monopolar as the default mode for general use, with bipolar and tripolar available as experimental or backup modes. Despite the broader current spread, monopolar’s combination of low current threshold, large dynamic range, and high stimulation efficiency outweighs its lower spatial selectivity for most patients. Channel-focused stimulation may help specific use cases (research, certain pitch-discrimination tasks) but has not produced demonstrable clinical speech-perception improvements in the populations studied.

Insertion geometry: lateral wall vs perimodiolar

The electrode array can be designed to sit either against the lateral wall of scala tympani (away from the modiolus, hugging the outer wall of the spiral) or perimodiolar (close to the modiolus, inside the spiral). The two designs have different current-spread profiles:

• Lateral-wall arrays (e.g., MED-EL Flex, Advanced Bionics SlimJ): straight or mildly curved electrodes that sit against the lateral cochlear wall. Easier to insert, gentler on cochlear structures, but the electrodes are 1.5–2 mm from the modiolar fibres they need to stimulate, requiring higher currents.
• Perimodiolar arrays (e.g., Cochlear Nucleus Contour Advance, Cochlear Slim Modiolar): pre-curved arrays designed to coil toward the modiolus during insertion. Electrodes sit closer to the spiral ganglion (0.5–1 mm), giving lower current thresholds and theoretically better current focusing, but require more complex insertion technique and may produce more cochlear trauma.

The clinical trade-off between the two designs has been studied extensively without producing a clear winner for routine implantation. Lateral-wall designs are typically preferred when preservation of residual hearing is a priority (electrode acoustic and hybrid implants); perimodiolar designs are typically preferred when maximum electrical efficiency and best-possible channel separation matter.

Insertion depth and the residual-hearing case

The cochlea’s apical region encodes low frequencies (200 Hz–1 kHz) where speech and music’s most informative pitch information lives. But the apical region is also where the array is hardest to reach — the cochlear spiral becomes tighter, the lumen narrower, and the risk of insertion trauma higher. Practical array insertion depths range from 18 mm (short-array hybrid devices intended to preserve low-frequency residual acoustic hearing) to 28–30 mm (full-coverage arrays intended to reach the apex).

The hybrid cochlear implant is a specific class designed for patients who have severe-to-profound high-frequency loss but useful low-frequency hearing remaining. The strategy: a short electrode array covers only the basal 18 mm of cochlea (encoding ~1.5 kHz and above electrically), leaving the apical low-frequency region undisturbed and acoustically stimulated by a co-fitted hearing aid in the same ear. This electric-acoustic stimulation (EAS) combination preserves the high-fidelity low-frequency hearing the patient already has while restoring high-frequency audibility electrically. Outcomes for hybrid users in candidacy-appropriate populations are excellent and often superior to standard-array CI alone for speech-in-noise.

Next lesson: the speech-processing strategies that turn the microphone’s acoustic signal into the per-electrode pulse trains delivered to this array.