The middle ear is a sophisticated and complex structure with a variety of functions, yet a delicate organ prone to injuries due to various reasons. Both, understanding and reconstructing its functions has always been an important topic for researchers from medical and technical background. Currently, human temporal bones are generally used as model for tests, experiments and validation of the numerical results. However, fresh human preparations are not always easily accessible and their mechanical properties vary with time and between individuals. Therefore we have built an anatomically based and functional middle ear model to serve as a reproducible test environment.
Our middle ear model was manufactured with the aid of 3D-printing technology. We have segmented the essential functional elements from micro computed tomography data (μCT) of a single temporal bone. The ossicles were 3D-printed by selective laser melting (SLM) and the soft tissues were casted with silicone rubber into 3D-printed molds. The ear canal, the tympanic cavity and the inner ear were artificially designed, but their design ensured the anatomically correct position of the tympanic membrane, ossicular ligaments and the oval window.
For the determination of their auditory properties we have conducted two kinds of tests: measurement of the stapes footplate response to sound and tympanometry of the model. Our experiments regarding the sound transmission showed that the model has a similar behavior to a human middle ear. The transfer function has a resonance frequency at around 1 kHz, the stapes' response is almost constant for frequencies below the resonance and a roll-off is observed above the resonance. The tympanometry results show that the compliance of the middle ear model is similar to the compliance of a healthy human middle ear.
We also present that we were able to manipulate the transmission behavior, so that healthy or pathological scenarios can be created. For this purpose we have built models with different mechanical properties by varying the hardness of the silicone rubber used for different structures, such as tympanic membrane, oval window and ossicle attachments in the range of Shore 10–40 A. This allowed us to set the transmission amplitudes in the plateau region higher, lower or within the tolerances of normal middle ears (Rosowski et al., 2007).
Our results showed that it is possible to build an artificial model of the human middle ear by using 3D-printing technology in combination with silicone rubber molding. We were able to reproduce the anatomical shape of the middle ear's essential elements with high accuracy and also assemble them into a functioning middle ear model. The acoustic behavior of the model can be reproduced and manipulated by the choice of material. If the issues such as resonance of the casing and steep roll-off slope in higher frequencies can be solved, this model creates a reproducible environment for experiments and can be useful for the evaluation of prosthetic devices.