Quasi-static transfer function of the rabbit middle ear' measured with a heterodyne interferometer with high-resolution position decoder.

Due to changes in ambient pressure and to the gas-exchange processes in the middle ear (ME) cavity, the ear is subject to ultra-low-frequency pressure variations, which are many orders of magnitude larger than the loudest acoustic pressures. Little quantitative data exist on how ME mechanics deals with these large quasi-static pressure changes and because of this lack of data, only few efforts could be made to incorporate quasi-static behavior into computer models. When designing and modeling ossicle prostheses and implantable ME hearing aids, the effects of large ossicle movements caused by quasi-static pressures should be taken into account. We investigated the response of the ME to slowly varying pressures by measuring the displacement of the umbo and the stapes in rabbit with a heterodyne interferometer with position decoder. Displacement versus pressure curves were obtained at linear pressure change rates between 200 Pa/s and 1.5 kPa/s, with amplitude +/-2.5 kPa. The change in stapes position associated with a pressure change is independent of pressure change rate (34 microm peak-to-peak at +/-2.5 kPa). The stapes displacement versus pressure curves are highly nonlinear and level off for pressures beyond +/-1 kPa. Stapes motion shows no measurable hysteresis at 1.5 kPa/s, which demonstrates that the annular ligament has little viscoelasticity. Hysteresis increases strongly at the lowest pressure change rates. The stapes moves in phase with the umbo and with pressure, but the sense of rotation of the hysteresis loop of stapes is phase inversed. Stapes motion is not a simple lever ratio mimic of umbo motion, but is the consequence of complex changes in ossicle joints and ossicle position. The change in umbo position produced by a +/-2.5 kPa pressure change decreases with increasing rate from 165 microm at 200 Pa/s to 118 microm at 1.5 kPa/s. Umbo motion already shows significant hysteresis at 1.5 kPa/s, but hysteresis increases further as pressure change rate decreases. We conclude that in the quasi-static regime, ossicle movement is not only governed by viscoelasticity, but that other effects become dominant as pressure change rate decreases below 1 kPa/s. The increasing hysteresis can be caused by increasing friction as speed of movement decreases, and incorporating speed-dependent friction coefficients will be essential to generate realistic models of ossicle movements at slow pressure change rates.