Direct bone conduction stimulation: effect of different transducer attachments

Abstract

Objective(s): When choosing a method to implant a transducer for direct bone conduction stimulation, several options are available with the trade-off being, among other aspects, between achieving a robust contact to the bone and keeping an open possibility for potential future explantation. The aim of this study is to compare how different attachment methods for direct bone conduction stimulation can affect the vibrations transmission to the cochleae. The ultimate goal is to gain more insights on the dynamical properties of the skull to apply them to the design and optimization of bone conduction devices for hearing rehabilitation. Study design: Experimental. Patients: Measurements were performed on four human heads. The tympanic membrane, the malleus and the incus were removed to expose the cochlear promontory on both sides of each subject. The subjects have not undergone any previous surgery as verified by visual inspection. Methods: Three different attachments are tested on eight sides: (A) flat small-sized surface, (B) flat wide surface and (C) two separated screws. The different typologies of contact to bone are established by three dummy implants and an adapter to attach the transducer giving a swept sine stimulus from 0.1 to 10 kHz. The response is evaluated in terms of cochlear promontory acceleration and ear canal sound pressure level (ECSP). The measurement setup consists of the following parts: human head, transducer with adaptor (to apply the stimulus), signal generator and analyzer (to drive the transducer and receive the recorded data), Laser Doppler Vibrometer (LDV, measuring the cochlear promontory acceleration), video to USB converter (to couple the built-in camera of LDV with the computer), microphones (to measure ECSP) and laptop (to save data). Results: Average results show slightly higher ECSP level and cochlear promontory acceleration for attachment A compared to attachment B especially at frequencies above 1 kHz. An improvement in transmission is achieved with attachment C compared to B in the frequency range 5–7 kHz, where the levels differ by nearly 10 dB both ipsi- and contralaterally. In the same frequency range, the transmission from attachment C appears to be slightly higher relative to attachment A as well, however no statistical significance is found. The comparison between attachment A and C is not straightforward due to high variability over frequencies and sometimes contradictory results between LDV and ECSP measurements. Conclusion: On an average level, the screw stimulation technique seems to improve the transmission at frequencies above 5 kHz. However,, when considering the whole frequency range, average results from the different attachment techniques are comparable. Still, conclusions for single subjects should be drawn with care as measurements show a very high inter-subject variability.