| An essential component in the process of hearing is the transformation of sound from acoustic to mechanical vibration in the middle ear. In order to study this phenomenon, computational models have been introduced to model the behavior of the tympanic membrane and its interaction with the surrounding acoustical spaces. Using such a computational model, one gains an increased understanding of the relationship between its structure and performance, which may assist in preventive, diagnostic, and reconstructive medical applications.; The goal of this present work is to expand the computational simulation capabilities of current eardrum models using modern finite element modeling techniques. A fully coupled structural acoustic model is presented using modern shell element technology. Anatomical geometries for the eardrum, acoustic spaces of the ear canal and middle ear cavity, and the ossicles are utilized via muCT imaging.; A new computational algorithm is used to compute the frequency response of this model over a wide frequency range. This approach uses the matrix Pade-via-Lanczos algorithm to construct reduced-order models around chosen reference frequencies, which can be solved efficiently at many frequencies within a frequency window. An adaptive algorithm is introduced to span a given frequency range by introducing new reference frequencies as necessary.; Results for the middle ear model, using this multifrequency algorithm, are presented for intact and modified middle ear anatomies. These modifications serve to demonstrate the utility of the computational approach in understanding the relationships between the morphological structure of the middle ear and its functionality. |
