High Resolution Computed Tomography Atlas of the Porcine Temporal Bone and Skull Base: Anatomical Correlates for Traumatic Brain Injury Research

Abstract

Brain injuries are a significant cause of morbidity and mortality worldwide. Auditory and vestibular dysfunction may occur following trauma to the temporal bone (TB), including the lateral skull base. The porcine model is a commonly used large animal model for investigating brain injury. Reports detailing porcine TB anatomy based on high resolution computed tomography (HRCT) imaging, however, are limited. Herein, we employ HRCT to evaluate and describe the bony anatomy of the porcine TB and lateral skull base. High-resolution multi-detector and cone beam CT were used to image porcine TBs (n = 16). TBs were analyzed for major anatomical structures and compared to human species. Porcine temporal bone anatomy was readily identifiable by HRCT. Although some variability exists, the ossicular chain, vestibule, cochlea, course of the facial nerve, and skull base are similar to those of humans. Major differences included position of the external auditory canal and mastoid, as well as presence of the petrous carotid canal. Study findings may serve as an atlas to evaluate the porcine middle and inner ear, as well as lateral skull base injuries for future porcine brain injury models or other studies that require CT-based analysis.

Introduction

Traumatic brain injuries (TBIs), are a critical public health concern in the United States and worldwide.^1–3 In the United States, there are >80,000 deaths and >5,000,000 patients with chronic neurological deficits that result from TBI.^1,2,4,5 Brain injury can result in loss of function in cognitive, sensory, perceptual, psychological, speech, and language capacities.^6,7 Classification of TBIs ranges from mild to moderate to severe based on the duration of loss of consciousness or mental status change, post-traumatic amnesia, and imaging results.⁸

Auditory and vestibular dysfunction has long been recognized as one of the possible consequences of brain injury, including temporal bone (TB) fracture.^9–13 The prevalence of audiovestibular problems following head injury is estimated to occur in up to 80% of patients.^8,14–17 Dating back to the descriptions of boxers with punch drunk syndrome in the 1920s,¹⁸ later known as “dementia pugilistica,”¹⁹ athletes involved in contact sports that sustain mild brain injury, such as concussions, have routinely indicated symptoms of audiovestibular dysfunction.^18,20,21 Moreover, military personnel also commonly report tinnitus and dizziness following brain injury.²²

Although numerous studies have analyzed human data following TBI, there are inherent limitations in retrospective analyses. Large animal models have emerged as a key method for investigating the effects and pathophysiological mechanisms of brain injury, as they allow serial and invasive experimentation.²³ In particular, there is a predilection for the use of large animal models with closed-head injury because of their capability to replicate tissue-level biomechanics.²⁴

Various animal models have been used to investigate the middle and inner ear, including mice,²⁵ rats,²⁶ guinea pigs,²⁷ sheep,^28,29 and cats.³⁰ The porcine TB has also been suggested as a potential model for investigating audiovestibular pathology,³¹ and a few studies have commented on porcine TB anatomy.^32–38 However, to date, there is limited description of porcine TB anatomy using high resolution computed tomography (HRCT), as prior studies provide only brief descriptions of CT-based anatomy.^32–34,38 Herein, we aim to evaluate and analyze porcine TB anatomy based on readily available high-resolution imaging.

Methods

Animal subjects

Common pig (Sus scrofa domestica) and Hanford miniature swine heads were included in the study. Both the common pig and Hanford miniature swine are routinely utilized in brain injury research. The common pig (n = 4 heads, 8 TB) were ∼5–7 months old at time of imaging. Hanford miniature swine (n = 4 heads, 8 TB) were 6 months old at time of imaging. The common pig heads were obtained from a local butcher (Cambridge, MA) and the Hanford miniature swine heads were procured from a research farm facility (Sinclair BioResources). Gender and weight at time of harvest were unknown. With regard to size, common pig and Hanford miniature swine reach sexual maturity between 3 and 5 months; however, they generally increase in size and weight after reaching maturity.³⁹

HRCT scans

Whole porcine heads were placed in a high resolution multi-detector CT scanner (Discovery 750 HD, General Electric, Milwaukee, WI). Images were acquired with 0.625 mm axial slices, a 0.2 mm slice gap, and 120 kVp and 240 mA. To visualize the ossicular chain, a cone beam CT (CBCT, 3D Accuitomo, J. Morito MFG Corp., Kyoto, Japan) of the specimen was performed following TB dissection (see next section, Porcine TB dissection for details). CBCT images were acquired at 90kV, 8mA, hi-resolution mode, exposure time 30.8 sec, field of view 60 × 60 mm, with a slice thickness of 0.5 mm. CBCT allows lower irradiation by trend⁴⁰ and higher accuracy because of the smaller voxel size,⁴¹ and is often used clinically for TB imaging of the middle ear and skull base.

Porcine TB dissection

In order to isolate the TBs for CBCT, the TBs were removed based on human TB techniques.⁴² In brief, the caudal portion of the skull was removed by an oscillating saw blade (Model 810, Stryker Corportation, MA) (Fig. 1A). Next, the brain and dura mater were removed from the skull base. Removal of dura revealed the middle cranial fossa and the superior surface of petrous bone, which was easily identified because of the bright white appearance and high-density characteristic of the pig TB (Fig. 1B,C). The TB was then isolated and removed using the autopsy saw with a circular cutter (1.5 inches or 38 mm inside diameter and 51 mm in length), (Fig. 1A). The circular cutter was positioned at an acute angle to generate a cone of tissue containing part of the external auditory canal (EAC), middle ear, and inner ear (Fig. 1D). The saw was taken down to soft tissue, which was then dissected free by an osteotome and heavy Mayo scissors. Muscular attachments to the styloid were typically the most medial aspect of the dissection. After the dissection, a TB HRCT scan was performed to exclude any sign of injury of the middle and inner ear in the TB plug, such as presence of fracture or ossicular chain disjunction (Fig. 2A, B).

FIG. 1.

Temporal bone extraction. (a) Autopsy saw and surgical instruments necessary for dissection of the porcine temporal bone. (b and c) Superior view of the skull base showing surface of petrous bone, box, and arrow. (d) Circular bone plug saw position for temporal bone dissection.

FIG. 2.

Computed tomography (CT) of the middle and inner ear after temporal bone dissection. A CT scan confirmed the integrity of the middle and inner ear after the dissection. (a) CT scan of the whole pig head before dissection. (b) CT scan of the temporal bone (TB) “plug” after dissection.

Porcine and human specimen comparisons

Qualitative and quantitative comparisons between porcine and human specimens were completed. To perform qualitative comparisons between normal human and porcine TB HRCT scan anatomy, the National Temporal Bone Hearing and Balance Pathology Resource Registry (Boston, MA) was used to identify normal human cases. All HRCT images were evaluated by a neuroradiologist (K.L.R.) and four otolaryngologists (R.M.K., S.R.B., A.K.R., E.D.K.) who have expertise in TB imaging and clinical anatomy.

To perform quantitative comparisons, DICOM data from the HRCT scans of the porcine heads were imported into OsiriX Lite v.9.0.2, in which measurements in the axial plane were obtained. Measurements included: maximal width and length of vestibule, canal lumen width and bony island width of the superior semicircular canal (SSCC) and lateral semicircular canal (LSCC), canal lumen width and inferior limb length of posterior semicircular canal (PSCC), vestibular aqueduct (VA) midpoint, and operculum. The mean values and standard deviation (SD) of the measurements were calculated, and then compared with normal historical human TB data as previously described.^43,44

Three-dimensional (3D) reconstructions

DICOM data from the HRCTs of the heads and TB plug were imported into 3D Slicer 4.8, where the bone was manually segmented into 3D surface meshes and exported as stereolithography (STL) files. Files were imported into 3D modeling software where physically based rendering (PBR) materials and lighting were applied for model production.

Results

Surface anatomy of the TB

The porcine skull base has a thick and compact appearance, with a generally smaller cranial vault than in humans. The gross anatomy of the TB is quite similar in terms of general morphology and position. Using 3D reconstructions from HRCT, the osseous parts of the temporal bone, including the squamous, mastoid, tympanic, and petrous portions, as well as the styloid process, can be visualized (Fig. 3B). The styloid process is longer and thicker than in humans (Fig. 3B–D). The tympanomastoid suture can be identified as well (Fig. 3D). The porcine EAC is long compared to that in humans, and is orientated in a posterosuperior to anteroinferior direction (Fig. 3B, D and Fig. 4). Unlike in humans, there is a significant amount of soft tissue covering the TBs laterally. Further, whereas the mastoid air cells in humans are found posteriorly to the tympanic cavity, in the pig they are oriented medial to the temporomandibular joint and anteroinferior to the tympanic cavity and EAC (Fig. 3C, Fig. 4A and Fig. 5A, C, D).

FIG. 3.

Three-dimensional (3D) reconstructions of the pig skull base. (a, a’) Superior View. (b) Left lateral view. (c) Inferior view. (d) Left posteromedial view.

FIG. 4.

Coronal computed tomography (CT) scan of a pig head from anterior to posterior (a–d), showing a long external auditory canal orientated strictly from posterosuperior to anteroinferior, and the relationship between the middle cranial fossa and the otic capsule. EAC, external auditory canal; MCF, middle cranial fossa; TM, tympanic membrane.

FIG. 5.

Comparison of human and pig temporal bone computed tomography (CT) anatomy. Axial CT scan of human (a) and pig (b) right temporal bones, and a coronal view of human (c) and pig (d) right temporal bones. LSCC, lateral semicircular canal; PSCC, posterior semicircular canal; SSCC, superior semicircular canal; IAC, internal auditory canal; EAC, external auditory canal; OW, oval window.

Tympanic cavity and ossicular chain

The tympanic cavity appears similar to that of humans with epitympanum, mesotympanum, and hypotympanum subdivisions. The ossicular chain (incus, malleus, and stapes) appears to have minor differences based on HRCT. The malleus is the most anterior ossicle. Its head articulates with the body of the incus at the malleoincudal joint (Fig. 5A, B, Fig. 6A, B, and Fig. 7D). The body of the porcine incus is smaller relative to the malleus in comparison with the human incus as previously described.^32,33 The long process of incus passes inferiorly and forms a right angle with the small lenticular process, which articulates with the stapes capitulum (incudostapedial joint) (Fig. 6A, B; Fig. 7F; Fig. 8G; and Fig. 9C, D). Ellenberger and coworkers⁴⁵ and Nickel and coworkers⁴⁶ described a fourth ossicle in the pig named “os lenticulare,” located between the stapes and incus; however, it could not be observed in HRCT images.

FIG. 6.

Three-dimensional (3D) reconstruction of the middle and inner ear structures from the cone beam computed tomography (CBCT) images. Superior view (A), medial wall view (B), and medial wall view without the ossicular chain (C). LSCC, lateral semicircular canal; PSCC, posterior semicircular canal; SSCC, superior semicircular canal.

FIG. 7.

Consecutive axial cone beam computed tomography (CBCT) images through the left temporal bone demonstrating the anatomy from cephalic to caudal. LSCC, lateral semicircular canal; PSCC, posterior semicircular canal; SSCC, superior semicircular canal; IAC, internal auditory canal; ICA, internal carotid artery; OW, oval window; RW, round window; EAC, external auditory canal; FN, facial nerve.

FIG. 8.

Consecutive coronal cone beam computed tomography (CBCT) images through the left temporal bone demonstrating the anatomy from anterior to posterior. LSCC, lateral semicircular canal; PSCC, posterior semicircular canal; SSCC, superior semicircular canal; IAC, internal auditory canal; OW, oval window; RW, round window; EAC, external auditory canal; FN, facial nerve; TM, tympanic membrane; SVN, superior vestibular nerve; IVN, inferior vestibular nerve.

FIG. 9.

Comparison of human and pig temporal bone plug computed tomography (CT) anatomy. Axial CT scan of human (a) and pig (b) left temporal bone plugs, and a coronal view of human (c) and pig (d) left temporal bone plugs. LSCC, lateral semicircular canal; SSCC, superior semicircular canal.

Prussak's space can be identified as the lateral epitympanic space with the scutum forming its lateral wall (Fig. 4B and Fig. 8E, F). The anterior wall of the tympanic cavity contains the eustachian tube and the tensor tympani attaching to the cochleariform process (Fig. 7H–J and Fig. 8E). The medial wall of the tympanic cavity separates the middle ear from the inner ear, and the lateral wall is composed of bone and the tympanic membrane as in humans (Fig. 4B and Fig. 8G). The posterior wall is delineated by part of the facial nerve inside the fallopian canal, as it forms the second genu below the oval window (Fig. 7I and Fig. 10E, F). The superior wall of the tympanic cavity is thick and includes contribution from the squamous portion of the TB (Fig. 4A). As mentioned previously, the mastoid air cells are anteroinferior to the tympanic cavity and EAC (Fig. 4A and Fig. 8A).

FIG. 10.

Axial Comparison of human and pig facial nerve course in the computed tomography (CT) scan. CT scan sequence from cephalic to caudal showing the facial nerve course and segments in human (a, c, e, g) and pig (b, d, f, h). IAC, internal auditory canal.

Facial nerve course

The internal auditory canal appears shorter than in humans (Fig. 5A, B), but the facial nerve follows a similar course. The facial nerve is anatomically similar to that in humans, having an anteriorly oriented labyrinthine segment with the first genu medial to and at the level of the tensor tympani (Fig. 7C; Fig. 8A–C; and Fig. 10A, B). The tympanic segment of the facial nerve runs in an anteroposterior direction in the fallopian canal between the first and second genu, showing the same inferior relationship to the lateral semicircular canal (Fig. 7D, E, G; Fig. 8G, H; and Fig. 10C, D). Following its second genu (Fig. 10E, F), the facial nerve courses inferiorly as the mastoid segment (Fig. 10G, H) before exiting from the temporal bone through the stylomastoid foramen situated between the mastoid and the styloid process (Fig. 3C and Fig 7K, L). The chorda tympani nerve could not be identified on HRCT.

Inner ear and skull base

The petrous bone is easily recognized on HRCT because of its high density. The inner ear, comprising the cochlea, vestibule, and semicircular canals, is invested by a thick otic capsule similar to that in humans (Fig. 6C). Also, similar to in humans, the superior semicircular canal is adjacent to dura. Anterior and superior, the middle cranial fossa overlying the temporal lobe appears to thicken as it projects with the squamous portion of the TB. However, unlike in humans, the porcine otic capsule is inferiorly (infratentorially) and posteriorly positioned in relationship to the middle fossa floor, and the lateral cerebellar folia overlies its superior margin (Fig. 3A and Fig. 4A). In humans, the temporal lobe is positioned over the otic capsule.

The cochlea had 3.5 turns compared with 2.5 turns in the human (Fig. 7I; Fig 8B–D; Fig. 11A, B). The cochlear aqueduct, which passes in the transverse plane inferior to the internal auditory canal, can be identified (Fig. 7J and Fig. 8I). The oval window with the stapes lying on top can be easily identified (Fig. 7F; Fig. 8G; and Fig. 9A, B), and the round window is located immediately inferior (Fig. 7I and Fig. 8H). The semicircular canals emerge from the superior, lateral, and posterior aspects of the vestibule (Figs. 5–8).

FIG. 11.

Comparison of human and pig cochlea computed tomography (CT) anatomy. Axial CT scan of human cochlea with 2.5 turns (a), and pig cochlea with 3.5 turns (b).

In comparison with the other inner ear structures and in humans, the vestibule and semicircular canals are small and thin (Table 1; Fig. 5; Fig. 9C, D). The vestibular aqueduct (Fig. 7E and Fig. 8K) can be seen from the posteromedial aspect of the vestibule, until it opens at the posterior aspect of the petrous ridge (Fig. 7D), and it generally appears to be larger than in humans (Table 1). CBCT allows identification of the singular nerve canal (Fig. 7H and Fig. 8H).

Table 1.

Measurement and Comparison of the Inner Ear Structures Based on HRCT

	Common pig (n = 8)		Hanford miniature swine (n = 8)		Human43,44
Structure	Mean	SD	Mean	SD	Mean	SD
Vestibule width	2.14	±0.07	2.12	±0.05	3.17	±0.22
Vestibule length	4.98	±0.35	4.84	±0.16	5.95	±0.22
Diameter of the IAC	2.14	±0.08	2.15	±0.07	3.7	±0.3
Length of the IAC	5.06	±0.14	4.83	±0.13	10.6	±1.2
SSCC canal lumen	0.67	±0.05	0.66	±0.03	1.31	±0.1
SSCC bony island width	4.64	±0.25	3.97	±0.18	5.06	±0.31
PSCC canal lumen	0.75	±0.18	0.61	±0.03	1.36	±0.11
PSCC inferior limb length	3.43	±0.32	3.21	±0.12	6.89	±0.61
LSCC canal lumen	0.56	±0.32	0.58	±0.03	1.38	±0.11
LSCC bony island width	2.88	±0.09	2.78	±0.14	3.64	±0.38
VA midpoint	0.96	±0.22	0.75	±0.02	0.4	±0
VA operculum	2.25	±0.40	1.52	±0.19	0.5	±0.1

The measures are presented in millimeters.

HRCT, high resolution computed tomography; SD, standard deviation; IAC, internal auditory canal; SSCC, superior semicircular canal; PSCC, posterior semicircular canal; LSCC, lateral semicircular canal, VA, vestibular aqueduct.

Major vessels of the TB

The sigmoid sinus appears smaller than in humans, coursing posterior to the mastoid air cells toward the jugular foramen (Fig. 3A, D; Fig. 5A, B; and Fig. 7J, L). The presence of a carotid canal in pigs is described in veterinary literature,^47,48 and appears to be medial to the eustachian tube; however, the petrous carotid canal was not readily identified on the axial CBCT images (Fig. 7H).

Discussion

This study provides detailed descriptions and analysis of porcine temporal bone anatomy based on HRCT. In our study the porcine TB anatomy was generally analogous the human TB with readily identifiable structures. The close anatomical similarities of the inner and middle ear on HRCT between the pig and the human are an advantage when considering it as an animal model to study audiovestibular dysfunction following TBI.

Prior studies have very briefly discussed CT-based structures of the porcine TB and skull base.^32–34,38 More specifically, the porcine TB model has been used to investigate inflammatory processes within the middle ear^34,49 and the morphology and biomechanics of the middle and inner ear,^31–33,50 and has been proposed as a model for surgical training in otology^32,35–37 because of its similarity to humans. Our article significantly builds on previous descriptions, providing technical and detailed descriptions of the middle and inner ear and skull base HRCT anatomy, which may be used as an atlas for future studies.

There are some notable anatomical differences between porcine and human TB. Principally, the position of the porcine EAC and mastoid air cells is different than in humans, which has also been identified in previous studies.^32,35,36,50 In an experimental study of human TB, Ilea and coworkers⁵¹ suggested that the mastoid portion of the TB might reduce the incidence of fracture in direct lateral trauma to the TB, by absorption and dispersion of kinetic energy. In addition, given the presence and location of mastoid air cells and the significant amount of soft tissue surrounding the TB laterally, there may be differences in how a lateral force is absorbed by a porcine TB, which may have correlations for brain injury research.

There are several limitations of the article worth highlighting. First, there is a small sample size of a specific age group, which did not allow for analysis based on specimen age. In humans, the development of the inner ear has been well described, and it reaches the adult size and configuration by the middle of the fetal gestational period (20–22 weeks).⁵² However, to the best of our knowledge, no study has definitely looked at the rate of growth and density of porcine TB over time. Further, although qualitative and quantitative analysis comparing our specimens is provided, our article was not designed to be a formal comparative anatomy study. Future studies may be able to rigorously address porcine anatomical osseous changes based on age and analogous comparison with humans. Indeed, a comparative anatomy study may ideally be completed through microdissection.

In summary, the current study provides a foundational HRCT atlas of the porcine anatomy illustrating major structures of the middle and inner ear, as well as the skull base, which may be used as a reference and future study.

Conclusion

This study provides technical and detailed descriptions of porcine TB anatomy based on HRCT. These findings may serve as a reference for evaluating the middle and inner ear, as well as lateral skull base injuries for future porcine brain injury models or other otology studies that necessitate HRCT imaging analysis.

Author Disclosure Statement

No competing financial interests exist.