The middle ear can be affected by a number of pathologies which may lead to hearing loss. To generate physiologically optimized middle ear prostheses, as well as a support for advanced middle ear surgery, the knowledge of the microanatomy of the middle ear is essential. In understanding the physiological conditions of the blood supply inside the tympanic cavity, complications in the middle ear surgery, for example, the necrosis of the long process of the
In the middle of the 16th century, the anatomists Andreas Vesalius (1514–1564) and Giovanni Filippo Ingrassia (1510–1580) discovered the auditory ossicles in the tympanic cavity of the human petrous portion. Ingrassia identified the
In the 20th century several scientists examined the three-dimensional (3D) structure of the human auditory ossicles with the following methods.
Oesterle [
G. T. Nager and M. Nager [
Hamberger et al. [
Anson and Winch [
In the attempt to perform a highly accurate simulation of the sound conduction system, a review of the literature revealed that there is no anatomical correct 3D model according to the state of technology. The results mentioned above, obtained by using cutting sections of decalcified tissue or separated auditory ossicles, are rather imprecise compared with today’s technological possibilities. New models are necessary for optimizing middle ear prosthesis [
The presented study developed a new, innovative, and more detailed microgrinding method of the human middle ear to produce a computer supported 3D model of the in situ positioned, anatomically correct ossicular chain in the tympanic cavity. The advantage of the technique is the high resolution of the model based on the very short intervals between the examined surfaces. But the time consuming pilot study had to be restricted to only one specimen in order to prove its superiority to the available 3D reconstructions in the literature. The attempt of the presented study is not to provide general statements on the microanatomy of the human middle ear but to introduce a very exact 3D model in high resolution based on histological sections and manual segmentation in terms of a proof of feasibility study. The analysis of the histology of the grinded planes revealed detailed information of the tissues involved, which were not apparent in the micro CT. As CT scans are probably the fastest way to gather data for 3D models or for finite element simulations, the informative value of this method has to be compared with CT based models. If the superiority of the 3D model based on the microgrinding can be demonstrated, an increase of man and machine power for detailed analysis of more samples of the human middle ear will be possible in the future.
A left human temporal bone was taken from a body donor (56-year-old man), who gave face-to-face and in free will his informed written consent (own testament during his lifetime) to bequeath his body for teaching and research at the university after his death within the body donation program of the Institute of Functional and Applied Anatomy, Hannover Medical School, Hannover, Germany. The sample was lesion-free and without pathologies. It was fixed in a mixture of 2.5% glutardialdehyde in 0.1 M sodium cacodylate buffer with pH 7.3 (Merck, Darmstadt, Germany) at 4°C for three days. Afterwards a dehydrating process was performed using a four-step ethanol series with increasing concentration, and the sample was then dried over night at 65°C in a drying cabinet. Subsequently, the fixed and dehydrated sample was embedded in uncolored epoxy resin (SpeciFix 20 Kit, Struers A/S, Rodoyre, Denmark) under vacuum conditions. To also fill the tympanic cavity with epoxy resin, a hole was pierced into the tympanic membrane with a needle. After drying, three holes were drilled vertically into the epoxy resin and filled with wooden sticks to serve as reference marks. To ensure the holes were drilled exactly vertically into the epoxy resin, this process was conducted at a lathe.
Additionally, one series of separated right ear auditory ossicles from the anatomic collection of the Department of Otolaryngology, Hannover Medical School, Germany, was used for
The epoxy resin embedded left human temporal bone as well the formalin-fixed separated ossicular chain of a right ear was examined with a
For the microgrinding, the embedded left human temporal bone was ground and polished with a grinding machine (Buehler beta with vector, with Lc Power Head, Buehler GMBH, Germany) and grinding paper P2500 (medium grain size 10
In every grinding run the abrasion of the sample was measured in three places near the reference marks by the thickness of the remaining sample with a digital micrometer caliper. Consequently, the abrasion of the sample in every grinding run was about 35
The surface of the sample was stained for two minutes with the modified method according to Mann-Dominici [
A total of 151 layers were ground and stained. The histological images were taken with a digital camera system (AxioCam MRc, Zeiss, Jena, Germany), attached to a microscope (Zeiss, SteREO Discovery.V20) in 4- to 100-fold magnification, and illuminated from an external cold light source. In total 20996 histological images were taken documenting the histological structures in different magnifications. With Adobe Photoshop (Adobe software Ireland Ltd.) every histological high magnification microscopic image of the fine structures was placed into the histological overview image of the same layer. Thus an image of the hard and soft tissues in high resolution together with orientation marks was obtained of every layer and saved as JPEG (.JPG).
For segmentation the data sets of the
The 3D model of the auditory ossicles created from the histological images was designed with the CAD program Rhinoceros 5 (64-bit; McNeel). Therefore the reference marks were replaced by rings at the CAD program surface (Figure
Working surface of the CAD program Rhinoceros 5. The markings of the reference marks are arranged in piles and colored in red, blue, and lilac. The adjustment lines for the microgrinding images are marked in white.
The generated model consists of the outer shape of the ossicles with its ligaments, muscles, the articular capsules of the ossicular joints, and the articulation cartilage on the joint surface. The inner structures of the ossicles, that is, the inner vascular system and cartilage areas, were also modeled. Therefore every histological image was loaded into the program and adjusted accordingly to the reference marks and then structures of interest were marked with lines. Finally, these lines were connected to each other and a freeform surface was generated.
The
In this study an embedded specimen of a left adult human temporal bone was examined by the microgrinding method with removal of circa 35
The histological images of the microgrinding surface as well the deeper transparent zones of the embedded middle ear show that not only the tympanic cavity but also the ossicular chains are covered by a thin layer of mucosa. These mucosal membranes are connected to each other by several thin plications, which contain the ligaments as well as blood vessels. The blood vessel system forms a fine network within the mucosal layer around the ossicles.
The major blood supply of the
Reflected light microscopic images of the grinding surface after surface staining with modified staining according to Mann-Dominici. (a) Inside the tympanic cavity the stained tympanic membrane (TM) and
The malleolar branch of the artery inside the anterior malleolar plication runs along the anterior process of the
Three blood vessels penetrate the
The main blood supply of this specimen concurs with the findings of G. T. Nager and M. Nager [
The blood supply of the
The blood supply of the
With the novel microgrinding technique blood vessels can be found not only in the mucosa around the ossicular chain but also inside the ossicles. Some nutrient vessels penetrate the bones of
The blood vessel system inside the
Chen et al. [
Nevertheless, by using uncolored epoxy resin and ground resin and grinding in very thin layers we can follow the course of the internal vessels of the ossicles through the nutrition pores into the vascular network inside the “mesenteries” of the tympanic cavity as well as in the mucosa covering the auditory ossicles.
By examining the grinding sections only one cartilaginous area can be found within the head of the
Oesterle [
Anson [
The appearance of marrow spaces and cartilage areas can be explained by the embryological development of the auditory ossicles. From the mesenchyme of the first branchial arch the head and neck of the
To expound possible supremacy of the time consuming grinding method, the information value was compared with a relatively easy and fast
3D models of the
The generated 3D model of the embedded specimen illustrates the ossicular chain in their anatomically correct orientation within the temporal cavity as well as their relation to each other. The nutrient openings in the bony surface of the ossicles can here clearly be seen. One
3D model of the surface of the
The volume of the whole ossicles and especially of the not calcified structures can be evaluated. The total volume of the
In comparison, the histological data set allows a differentiated analysis of all cellular and extracellular structures of the ossicles as well as of the adjacent ligaments and articular structures in between. After reconstructing the auditory ossicles from histological images based on the microgrinding technique, similar volumes of the same ossicles can be estimated (i.e.,
Compared with the not calcified parts in the
In a study from Rau et al. [
Anson and Winch [
The 3D model of the intraosseous blood vessel system of
Our study shows that the microgrinding method is necessary to get all anatomical characteristics of the human auditory ossicles. By comparing the histological images of the stained plane with a parallel look to the deeper zones of the specimen by using uncolored epoxy resin, the 3D reconstruction is possible in a very high resolution. For designing a 3D model of the middle ear structures, the distance between the planes has to measure about 34
Whereas the time consuming microgrinding method and the high amount of workload in the adjustment process within the CAD program are necessary, the benefit becomes visible when compared with the results extracted from the
In contrast, the 3D model of the microgrinding images differentiates the histoanatomical structures and can rebuild, for example, the branched blood vessel system and the cartilage areas as well the course of collagen fibers. In order to substantiate these findings and for analyzing variations in anatomy, a second study with more specimens has to be undertaken.
An anatomically correct 3D model of the ossicular chain inside the intact middle ear is essential for the construction of physically optimized middle ear prostheses and for advanced middle ear surgery. In addition, it can be also used for computer based mechanical simulation of the sound transmission system.
All authors proclaim that there are no conflicts of interest with regard to research, authorship, content, and/or publication of this article.
Nils Prenzler and Gudrun Brandes contributed equally to this work.
The project was funded by the German Research Foundation (DFG, Collaborative Research Centre SFB 599, subproject D1) and profited from interactions within the Cluster of Excellence Hearing4All.
The authors appreciate the constructive discussions with Peter Behrens (Cluster of Excellence “Hearing4all”, Institute of Inorganic Chemistry, Leibniz Universität Hannover, Hannover, Germany) and Peter Müller (Helmholtz Centre for Infection Research, Braunschweig, Germany). Also the authors want to acknowledge the Institute of Functional and Applied Anatomy, MHH, and Peter Erfurt (NIFE, Hannover, Germany) for the preparation of the temporal bone.