Ultrasonic Cochlear Excitation — Device Scoping: where the proposal lands in the functional chain, what it would have to be, and what would have to be shown

A falsification test says whether an idea is wrong; it does not say what would be built if it were right. This scoping document fills that gap for acoustic rather than electrical stimulation of the cochlea: functional placement within an eight-stage decomposition of hearing assistance, the acoustic output envelope imposed by IEC 60601-2-37 and FDA Track 3 limits, a reverse-engineered power budget, the standards that would apply, and the staged evidence required before any human exposure.

Three quantitative findings shape it. The binding regulatory constraint is the time-average intensity limit rather than the mechanical index, because a hearing device runs continuously — at the pulse-average ceiling the permitted duty cycle is 0.38%. An eight-channel aperture operating at that ceiling draws about 1.1 W against a cochlear implant's 30 mW, so a wearable version would require the biological excitation threshold to sit near one percent of the diagnostic ceiling. And an acoustic aperture is bidirectional, so the same hardware that stimulates can image what it stimulates.

Three architectures are evaluated — a steered focus, ultrasonically powered implanted motes, and a passive graded resonator array — and each is closed or constrained for slow-growth retrocochlear loss for a different reason: nerve dependency, artifact burden on the surveillance imaging that defines this population's care, and duplication of a cochlea that here is intact. Because every closure is population-specific, a cross-population section states the specification that survives for each, with its first gate. Regulatory path, thermal limit, power budget and imaging burden independently select the same product: a clinic-sited instrument used for minutes, not a worn device.

A safety question with no direct evidence either way — the effect of repeated ultrasonic exposure on schwannoma proliferation, where the literature reports both inhibition and enhancement in different cell types — is stated as a gate rather than assumed benign.

Preprint, not peer reviewed. Prior-art searching was targeted and conducted at abstract and extract level. A declaration page tags every load-bearing number by provenance — 21 sourced, 4 unverified, 9 assumed, 4 computed — and separates decision-relevant quantities from design-relevant ones. One published dimension used here is flagged as internally inconsistent with its own reported results. 

Keywords. focused ultrasound; cochlear implant; acoustic output limits; IEC 60601-2-37; mechanical index; spatialpeak
temporal-average intensity; ultrasonically powered implant; artificial basilar membrane; MEMS resonator array;
bone-conducted ultrasound; medical device scoping; vestibular schwannoma; ALARA