We have recently developed an implantable piezoelectric hearing aid transdu
cer that is suitable for implantation in patients with sensorineural hearin
g loss. The transducer does not transmit sound but conducts micromechanical
vibrations to the cochlea. In ten cat ears we investigated the efficiency
of the implantable transducer with respect to the direct transfer of vibrat
ions within the audible frequency range via the ossicles to the cochlea or
directly into the vestibule. The acoustically evoked brainstem potential (A
BR) threshold was determined prior to implantation, and the middle ear was
then opened and the piezoelectric transducer coupled to the ossicles or to
the perilymph. Acoustically evoked brainstem potentials were recorded follo
wing stimulation at the umbo, long process of the incus, stapes head, stape
s foot plate, and in the vestibulum. Comparisons of the acoustically and me
chanically evoked thresholds revealed a good con-elation of the two stimula
tion levels. An electrical transducer voltage of 1 V-RMS produced equivalen
t sound pressure levels (SPL) of 100-128 dB at the tympanic membrane. To as
sess the hearing we compared stimulus-dependent latencies of the early pote
ntials (peaks P1-P5) and thresholds. This evaluation wets based on four ear
s with normal hearing in which the piezoelectric transducer was coupled to
the long process of the incus. The mean values of the latencies and their s
cattering range correlated extremely well in the two stimulation modes. The
y were nearly identical when the equivalent SPL of 100 dB was assigned to t
he maximally applied electrical level of 0 dB. These in vitro and in vivo f
indings demonstrate that the characteristics of the transducer warrant its
development further from the prototype stage to become a component of an im
plantable hearing device for patients with sensorineural hearing loss.