Accessible and Informative Sectioned Images, Color-Coded Images, and Surface Models of the Ear

Accessible and Informative SectionedImages, Color-Coded Images, and

Surface Models of the EarHYO SEOK PARK,1 MIN SUK CHUNG,2 DONG SUN SHIN,2 YONG WOOK JUNG,1

AND JIN SEO PARK1*1Department of Anatomy, Dongguk University College of Medicine, Gyeongju,

Republic of Korea2Department of Anatomy, Ajou University School of Medicine, Suwon, Republic of Korea

ABSTRACTIn our previous research, we created state-of-the-art sectioned

images, color-coded images, and surface models of the human ear. Ourear data would be more beneficial and informative if they were more eas-ily accessible. Therefore, the purpose of this study was to distribute thebrowsing software and the PDF file in which ear images are to be readilyobtainable and freely explored. Another goal was to inform otherresearchers of our methods for establishing the browsing software andthe PDF file. To achieve this, sectioned images and color-coded images ofear were prepared (voxel size 0.1 mm). In the color-coded images, struc-tures related to hearing, equilibrium, and structures originated from thefirst and second pharyngeal arches were segmented supplementarily. Thesectioned and color-coded images of right ear were added to the browsingsoftware, which displayed the images serially along with structurenames. The surface models were reconstructed to be combined into thePDF file where they could be freely manipulated. Using the browsingsoftware and PDF file, sectional and three-dimensional shapes of earstructures could be comprehended in detail. Furthermore, using the PDFfile, clinical knowledge could be identified through virtual otoscopy.Therefore, the presented educational tools will be helpful to medical stu-dents and otologists by improving their knowledge of ear anatomy. Thebrowsing software and PDF file can be downloaded without charge andregistration at our homepage (http://anatomy.dongguk.ac.kr/ear/). AnatRec, 296:1180–1186, 2013. VC 2013 Wiley Periodicals, Inc.

Key words: ear; anatomy; cross-sectional anatomy;three-dimensional imaging; hearing; posturalequilibrium; pharyngeal arch; visible humanprojects

Grant sponsor: National Research Foundation of Korea(NRF); Grant number: 2010-0023360.

*Correspondence to: Jin Seo Park, Department of Anatomy,Dongguk University College of Medicine, 87 Dongdae-ro,Gyeongju, Republic of Korea 780-350. Fax: 182-54-770-2402.E-mail: park93@dongguk.ac.kr

Received 21 October 2012; Accepted 18 April 2013.

DOI 10.1002/ar.22719Published online 27 May 2013 in Wiley Online Library(wileyonlinelibrary.com).

THE ANATOMICAL RECORD 296:1180–1186 (2013)

VVC 2013 WILEY PERIODICALS, INC.

INTRODUCTION

Human ear structures are hardly accessible by bothcadaver dissection and microscopy; most learning toolsfor displaying the complex anatomy of the ear have beenlimited to plastic models for several decades. To escapethis limitation, researchers have developed new learningtools, consisting of three-dimensional (3D) models (Wanget al., 2006, 2007; Li et al., 2007; Sorensen et al., 2009;Phillips et al., 2012). For that, basic image data wereobtained by anatomists or clinicians, while softwarepackages, on which the processed images are explored,were developed by computer engineers.

Our Visible Korean project team produced the excep-tional ear data as follows. In sectioned images of acadaver head (Park et al., 2009, 2010b), ear images(voxel size, 0.1 3 0.1 3 0.1 mm3; color depth, 48-bitcolor) were selected, and color-coded images displaying31 segmented ear structures were elaborated. Based onthe color-coded images, surface models were built onMaya version 2009 (Autodesk, Inc., San Rafael, CA).From this ear data, both macroscopic (e.g. cochlear ductand semicircular duct) and microscopic structures (e.g.tensor tympani muscle, stapedius muscle, and chordatympani) could be observed in detail (Jang et al., 2011).However, our ear data were not easily accessible becauseapplication software was not presented. Namely, the sec-tioned images could not be browsed along with accompa-nying color-coded images; the surface models could beoperated only on Maya, which is very expensive andcomplicated software. Therefore, we tried that our datawere applied to ear software of other research groupsuch as visible ear of Massachusetts Eye and Ear Infir-mary (http://otopathologynetwork.org) and A Brown andHerbranson Imaging Company (http://ehuman.com); butwe did not success because they were completionsoftware, not allowing modification or supplement bynot-computer engineer.

Meanwhile, we have already released browsing soft-ware in which the common user can easily access andbrowse sectioned images of the complete male body (Shinet al., 2011a) and a cadaver head (Shin et al., 2012c).Further, we have demonstrated that the free softwarepackage Adobe Reader Windows version 9 (Adobe Sys-tems, Inc., San Jose, CA) can display surface models ofthe complete male body and a cadaver head in a portabledocument format (PDF) file (Shin et al., 2012b,c).

The purpose of this research was to distribute thebrowsing software and the PDF file of ear, on which sec-tioned images, color-coded images, and surface models ofdetailed ear structures can be accurately explored,thereby making ear anatomy easier to learn. Anothergoal was to inform other researchers of our methods forestablishing the applications; eventually to enable themedical experts to better utilize their own specialized2D and 3D data. To achieve these goals, the sectionedimages and color-coded images of the right ear wereupdated, and those images and surface models were putinto the browsing software and PDF file, respectively.

MATERIALS AND METHODS

In a previous study, our research team producedsectioned images of a cadaver head (number, 2,341; reso-lution, 4,368 3 2,912; voxel size, 0.1 mm; color depth,48-bit color; file format, tagged image file format (TIFF))(Park et al., 2009, 2010b). From those, sectioned imagesof the ear region were selected (number, 221; resolution,1,622 3 420; voxel size, 0.1 mm), then primary color-coded images of 31 ear structures were produced (num-ber, 221; resolution, 1,622 3 420; voxel size, 0.1 mm) onPhotoshop CS5 version 12 (Adobe Systems, Inc., SanJose, CA) (Table 1; Jang et al., 2011).

In the sectioned and primary color-coded images ofear, the left half was cut off (resolution, 811 3 420)(Figs. 1A,B and 2A). In the primary color-coded imagesof the right ear, hearing and equilibrium structureswere colored blue and red, respectively. Except for theblue and red structures, chroma of the remnant struc-tures was reduced to create the secondary color-codedimages (Figs. 1C and 2A; Table 1). In the primary color-coded images, structures originating from the first andsecond pharyngeal arches in the embryo stage were col-ored (first pharyngeal arch, blue; second pharyngealarch, red) to prepare the tertiary color-coded images(Figs. 1D and 2A; Table 1).

Margins in all images, beyond for the middle and in-ternal ears, were simultaneously cut off to obtain thecropped ear (resolution, 406 3 210) (Fig. 2B).

Previously, browsing software was developed in orderto explore the sectioned and color-coded images ofcadaver head using the C# language of Microsoft VisualStudio. NET 2003 (Microsoft Corporation, Redmond,

TABLE 1. Thirty-one structures of the right ear segmented in sectioned images and reconstructed to buildsurface models

Region Structures

External ear Skin, external acoustic meatus, tympanic membranea

Middle ear Tympanic cavity, oval windowa, stapesa,b, incusa,c, malleusa,c, tensor tympani musclec,stapedius muscleb, auditory tube

Internal ear Anterior semicircular canald,e, posterior semicircular canald,e, lateral semicircular canald,e,cochleaa,e, internal acoustic meatus, utricled, sacculed, anterior semicircular ductd,posterior semicircular ductd, lateral semicircular ductd, cochlear ducta

Other Occipital bone, temporal bone, internal carotid artery, superior petrosal sinus, brainstem,facial nerve, chorda tympani, vestibular nerved, cochlear nervea

aStructures for hearing.bStructures for second pharyngeal arch.cStructures for first pharyngeal archdStructures for equilibrium.eStructure for which the surface models were not reconstructed.

ACCESSIBLE AND INFORMATIVE IMAGES OF EAR 1181

WA) (Shin et al., 2011a, 2012c). The software was com-posed of operating and image files that were replacedwith our new data: not cropped ear and cropped ear insequence. The operating and new image files were trans-formed into Browsing[lowem]software[lowem](Male[lowe-m]ear).exe, which was installation file, on the NullsoftScriptable Install System of NSIS Media.

The ear surface models were built from the primarycolor-coded images of bilateral ears and saved as stereo-lithography (STL) files on Maya (Jang et al., 2011). OnMaya, the surface models of bilateral ears were dividedinto left and right structures except brainstem and occi-pital bone. Using the Right Hemisphere Deep Explora-tion Standard (San Ramon, CA), STL files were

Fig. 1. Sectioned and color-coded images of the right ear which wereviewed from below. Sectioned images (A; row), primary color-codedimages (B; row), secondary color-coded images (C; row), and tertiarycolor-coded images (D; row) in upper (E; column) and lower (F; column)

levels around the middle ear. In the secondary color-coded images (C;row), blue and red colors signify hearing and equilibrium structures,respectively. In the tertiary color-coded images (D; row), blue and redcolors originate from first and second pharyngeal arches in that order.

1182 PARK ET AL.

categorized into right and left parts: external ear (right),external ear (left), middle ear (right), middle ear (left),internal ear (right), internal ear (left), other (right),other (left), and other (bilateral). In each part, the sur-face models were arranged in official anatomical terms(Table 1; FCAT, 1998). After finishing the coordination,all models in the STL files were gathered and saved asMale[lowem]ear.pdf using the 3D Reviewer, accompany-ing software of Acrobat 9.0 Pro Extended (Adobe Sys-tems, Inc., San Jose, CA) (Shin et al., 2012b,c). Whenthe PDF file was opened on Adobe Reader, the anatomi-cal terms of the parts were displayed in the model treewindow (Fig. 2C).

To learn anatomy, it was advantageous that modelswere assembled according to region. Thus, each regionwas assembled to be in bookmark window of the PDFfile on Acrobat as follows: bilateral ear (anterosuperiorview), right ear, right ear without temporal bone, hear-ing ear, equilibrium ear, first pharyngeal arch ear, andsecond pharyngeal arch ear (Fig. 2C; Table 1).

A previous article published in the Anatomical Recordjournal explaining the sectioned images and surfacemodels of the ear (Jang et al., 2011) was attached in thePDF file. Page 1 was already occupied by ear surfacemodels, so the article was included in the remainingpages.

RESULTS

The browsing software (253 MB) and PDF file (50.5MB) can be downloaded without charge or registrationat the homepage of the Department of Anatomy, Dong-guk University College of Medicine (http://anatomy.dong-guk.ac.kr/ear/).

In the browsing software, only right ear could beshown because it was programmed based on the rightside. In the PDF file, right, left, or bilateral structurescould be shown like bookmark window because it wasbuilt from images of bilateral ears.

After installing the browsing software by one-clickingthe EXE file, the sectioned images along with three setsof accompanying color-coded images could be easily

browsed in real-time. The user was able to select theimages either by using the scroll bar or by typing animage number into the software. The neighboringimages were continuously displayed by clicking the toolsin the software. When the user placed the mouse pointerover a structure in the sectioned or color-coded images,its name appeared as a tool tip text (Fig. 2A,B).

Next, ear surface models in the PDF file wereexplored on Adobe Reader version 9 and higher. The sur-face models could be suitably zoomed-in or zoomed-out,as well as freely rotated and shifted using mouse dragwith its buttons pushed. The surface models could bemade semi-transparent in order to view the back struc-tures of the models. When the user clicked a model, itsstructure name was highlighted in the model tree win-dow. Clicking structure names in the model tree windowprompted the appearance of matching surface models ei-ther individually or by group (Fig. 2C).

Ear anatomy could be easily comprehended by observ-ing sectional planes on the browsing software and 3Dshape on the PDF file. Examples are as follows.

In the sectioned and primary color-coded images onbrowsing software, the incus was located between themalleus and the stapes. The stapes fitted into an ovalwindow at the medial wall of the tympanic cavity, and itwas connected with the cochlea through the oval win-dow. In a sectioned image, a few holes of the spiral coch-lea were shown. By these considerations, the user couldunderstand the sectional shape of the hearing structures(Figs. 1A,B and 3A,B). For more convenience, the struc-tures related to hearing could be verified independentlyin the secondary color-coded images (Fig. 1C). The terti-ary color-coded images were supplementary to help com-prehend the embryonic origins of these structures.

In the PDF file, the tympanic membrane displayedconcavity toward the external acoustic meatus as thetensor tympani muscle medially pulled the tympanicmembrane through the embedded the malleus handle.The rounded superior head of malleus was articulatedwith the large body of incus. The long limb of incus liedparallel to the malleus handle, and its interior endencountered with the stapes. The spiral cochlea made

Fig. 2. Browsing software and PDF file of the right ear. Using the browsing software, sectioned imagesalong with accompanying primary, secondary, and tertiary color-coded images of the not cropped ear (A)as well as the cropped ear (B) are shown from below. Furthermore, the name of each structure is dis-played when the mouse pointer floats over a specific structure (A and B). PDF file is composed of amodel tree window, bookmark window, and surface models window (C).

ACCESSIBLE AND INFORMATIVE IMAGES OF EAR 1183

2.5 turns around a bony core, where cochlear nerveanchored. The large basal turn of the cochlea producedthe promontory of the tympanic cavity (Fig. 3C).

Using the browsing software, the holes of the anteriorand posterior semicircular ducts were observed.Although the user could expect that the ducts wereright-angled, it was unclear in a sectional view (Fig. 4A).In the PDF file, the user could accurately observe thatthe two ducts formed a right angle with the commonmembranous crus as the center. The lateral semicircularduct was smaller than other ducts (Fig. 4B,C).

In the PDF file, the tympanic membrane was observedthrough virtual otoscopy. The right-sided membrane wasdivided into four quadrants: anterosuperior (I), anteroin-ferior (II), posteroinferior (III), and posterosuperior (IV).The malleus handle was criteria intervening betweenquadrants I and IV, and the umbo was meeting point offour quadrants. The quadrants are clinically importantas they are used to describe the locations of lesions. Forexample, quadrant II shows the location of “cone oflight,” which is helpful in determining the tension of thetympanic membrane (Fig. 5A). In succession, virtualtympanoplasty was performed and the results comparedwith those of real surgery. After removing tympanicmembrane, in the superior part of quadrant IV, long

limb of the incus and chorda tympani could be seen. Thechorda tympani, originating from the facial nerve,crossed between the malleus and incus (Fig. 5A). Theanatomical and clinical knowledge in the PDF file wereidentical to that of a textbook (Fig. 5B; Clemente, 2010).

DISCUSSION

In our previous research, we created state-of-the-artscientific sectioned images as well as surface models ofthe ear (Jang et al., 2009), whole body (Park et al.,2005a,b, 2007, 2008; Shin et al., 2009a,b, 2011b,c,2012d), head (Park et al., 2009, 2010a,b; Shin et al.,2012a), and pelvis (Hwang et al., 2009). We expectedthat our data could be applied by computer engineers,but this did not happen; as a result, our data were notinformative since there was no easily accessible applica-tion developed. Therefore, we tried to create the prod-ucts of our data by ourselves; finally in this research, wepresented the outcomes which consisted of the browsingsoftware for sectioned images and the PDF file for sur-face models of ear.

Our developed browsing software and PDF file of earare easily obtainable. On the installed browsing soft-ware, the sectioned images along with accompanying

Fig. 3. Sectioned image, primary color-coded image, and surface models of right middle and internalears with annotations. Real sectional shapes of structures around ear ossicles are shown from below insectioned images (A), and positional relationships of the structures can be comprehended based on colorin the primary color-coded images (B). Real three-dimensional shape of ear ossicles and cochlea areshown (C).

Fig. 4. Semicircular ducts in sectioned images and surface models of the right ear. Anterior and poste-rior semicircular ducts are at right angles in a sectioned image (A), superior view of right ear surface mod-els (B), and lateral view of left ear surface models (C), as indicated by the red line and arrow.

1184 PARK ET AL.

color-coded images can be conveniently browsed usingthe mouse or keyboard, and names of the structures canbe displayed by placing the mouse pointer over theimages (Fig. 2A,B). Further, the PDF file can be viewedfor free on Adobe Reader. In the PDF file, surface modelsof each structure can be selected, rotated, zoomed-in,and zoomed-out freely using the mouse (Figs. 2C, 3C,4B,C, and 5A).

Users must be able to easily access the applications,as mentioned above; however, more importantly, accu-rate knowledge of anatomy should be obtainable fromthe informative applications. The sectioned images dis-play accurate anatomical information as they were madedirectly from the human body. The color-coded imagesand surface models are also precise, as they were cre-ated based on the sectioned images. In the browsingsoftware and PDF file, not only the sectional plane butalso the 3D shape of each structure can be observed andnames of the structures can be displayed as they werebuilt form the sectioned and color-coded images (Figs. 2–4). Therefore, users can acquire literal anatomicalknowledge of the ear using our applications. In clinics,an otologist has to interpret the computed tomographs ofhead to identify the ear components (Stimmer, 2011;Phillips et al., 2012). The browsing software and PDFfile could be utilized as the reference images.

Our applications can meet the demands of various sci-entific fields through upgraded or altered color-codedimages and surface models. In ear physiology, structuresrelated to hearing and equilibrium are very important.Thus, these structures were included in the secondarycolor-coded images and a bookmark in the PDF file. Inear embryology, structures originating from the first andsecond pharyngeal arches were involved in the tertiarycolor-coded images and another bookmark also (Figs. 1and 2). A textbook explanation about the ear could beinserted in the PDF file. Likewise, we hope that users orother researchers will promptly build the browsing soft-ware and PDF file according to their purposes

To prove the objectivity of our surface models, we per-formed virtual otoscopy and tympanoplasty. As a result,our PDF file could be applied as training tool in otology

as follows: using the PDF file, an otologist sees the tym-panic membrane with its folds in quadrant IV and he/she virtually removes the tympanic membrane. Thedoctor verifies that the malleus handle as well aschorda tympani are located inside of tympanic mem-brane’s folds. In the four quadrants of the membrane, itis proved that quadrant II is safest zone because no oneis very close inside the quadrant II. In addition, thedoctor can observe other structures in the middle andinternal ears (Fig. 5A). If it is real otoscopy, the tym-panic membrane and the some folds can be onlyobserved, whereas the chorda tympani and its relatedstructures behind the tympanic membrane are notobserved. It is difficult to realize the location of struc-tures behind the tympanic membrane in quadrant IVand II (Fig. 5B). Without virtual otoscopy and tympano-plasty, the resident is trained by the recorded real prac-tice movies or illustrations that are not interactive andstereoscopic (Marchioni et al., 2011). Therefore, in away, virtual otoscopy and tympanoplasty using the PDFfile can be more informative although the data are notyet sufficient for use in a clinical setting. Accordingly,we hope that the virtual simulation is improved byother researchers not only using our scientific data butalso adding other objective contents such as visible earof Massachusetts Eye and Ear Infirmary (http://otopathologynetwork.org).

Junior doctors of otology as well as most medical stu-dents find ear anatomy difficult when only non-interac-tive tools such as textbooks, plastic models, andrecorded dissection movies are available for learning.They feel uncomfortable with otoscopy and tympano-plasty of patients because it is known how difficult anddangerous the practices are. Our applications will beable to solve somewhat the uncomfortable feelingbecause most ear structures can be observed in detailwith real color using personal computer-based, off-line,interactive environment in our applications (Figs. 1–5);furthermore, they can be accessed and used easily freeof charge. Our developed applications can be easilyupgraded according to the needs of researchers. Thebrowsing software and PDF file will be helpful to

Fig. 5. Virtual and real otoscopy of the right ear. After the tympanic membrane is removed and opacityof temporal bone is adjusted semitranslucently in the PDF file, the membrane is divided virtually into fourquadrants. Shape of ear structures can be verified regardless of their locations (A). At the same location,only the tympanic membrane and some folds can be observed by real otoscopy, presented by a textbook(B; Clemente, 2010).

ACCESSIBLE AND INFORMATIVE IMAGES OF EAR 1185

medical students and otologists by improving theirknowledge of ear anatomy.

LITERATURE CITED

Clemente CD. 2009. Anatomy: a regional atlas of the human body.6th ed. Baltimore, Philadelphia, PA: Lippincott Williams &Wilkins. p 660.

Federative Committee on Anatomical Terminology (FCAT). 1998.Terminologia anatomica: international anatomical terminology.New York: Thieme

Jang HG, Chung MS, Shin DS, Park SK, Cheon KS, Park HS, Park JS.2011. Segmentation and surface reconstruction of the detailed earstructures, identified in sectioned images. Anat Rec 294:559–564.

Li PM, Wang H, Northrop C, Merchant SN, Nadol JB, Jr. 2007.Anatomy of the round window and hook region of the cochleawith implications for cochlear implantation and other endoco-chlear surgical procedures. Otol Neurotol. 28:641–648.

Marchioni D, Molteni G, Presutti L. 2011. Endoscopic anatomy ofthe middle ear. Indian J Otolaryngol Head Neck Surg 63:101–113.

Hwang SB, Chung MS, Hwang YI, Park HS, Har DH, Shin DS,Shin BS, Park JS. 2010. Improved sectioned images of the femalepelvis showing detailed urogenital and neighboring structures.Korean J Phys Anthropol 23:187–198.

Park JS, Chung MS, Chi J-G, Park HS, Shin DS. 2010a. Segmenta-tion of cerebral gyri in the sectioned images by referring to vol-ume model. J Korean Med Sci 25:1710–1715.

Park JS, Chung MS, Hwang SB, Lee YS, Har DH, Park HS. 2005a.Technical report on semiautomatic segmentation by using theAdobe Photoshop. J Digit Imaging 18:333–343.

Park JS, Chung MS, Hwang SB, Lee YS, Har DH, Park HS. 2005b.Visible Korean Human. Improved serially sectioned images of theentire body. IEEE Trans Med Imaging 24:352–360.

Park JS, Chung MS, Park HS, Shin DS, Har DH, Cho ZH, Kim YB,Han JY, Chi JG. 2010b. A proposal of new reference system forthe standard axial, sagittal, coronal planes of brain based on theserially-sectioned images. J Korean Med Sci 25:135–141.

Park JS, Chung MS, Shin DS, Har D-H, Cho ZH, Kim YB, Han JY,Chi JG. 2009. Sectioned images of the cadaver head including thebrain and correspondences with ultrahigh field 7.0 T MRIs. ProcIEEE 97:1988–1996.

Park JS, Jung YW, Lee JW, Shin DS, Chung MS, Riemer M, Han-dels H. 2008. Generating useful images for medical applicationsfrom the Visible Korean Human. Comput Methods ProgramsBiomed 92:257–266.

Park JS, Shin DS, Chung MS, Hwang SB, Chung J. 2007. Tech-nique of semiautomatic surface reconstruction of the Visible

Korean Human data using commercial software. Clin Anat20:871–879.

Phillips GS, LoGerfo SE, Richardson ML, Anzai Y. 2012. InteractiveWeb-based learning module on CT of the temporal bone: anatomyand pathology. Radiographics 32:E85-E105.

Shin DS, Chung MS, Lee JW, Park JS, Chung J, Lee SB, Lee SH.2009a. Advanced surface reconstruction technique to builddetailed surface models of liver and neighboring structures fromthe Visible Korean Human. J Korean Med Sci 24:375–383.

Shin DS, Chung MS, Park HS, Park JS, Hwang SB. 2011a. Brows-ing software of the Visible Korean data used for teaching sec-tional anatomy. Anat Sci Educ 4:327–332.

Shin DS, Chung MS, Park JS, Park HS, Lee SB, Lee SH, Choi HN,Riemer M, Handels H, Lee JE, Jung W. 2011b. Three-dimensionalsurface models of detailed lumbosacral structures reconstructedfrom the Visible Korean. Anal Anat 193:64–70.

Shin DS, Chung MS, Park JS. 2012a. Systematized methods of sur-face reconstruction from the serial sectioned images of a cadaverhead. J Craniofac Surg 23:190–194.

Shin DS, Chung MS, Park JS, Park HS, Lee S, Moon YL, Jang HG.2012b. Portable document format file showing the surface modelsof cadaver whole body. J Korean Med Sci 27:849–856.

Shin DS, Jang HG, Park JS, Park HS, Lee S, Chung MS. 2012c. Ac-cessible and informative sectioned images and surface models of acadaver head. J Craniofac Surg 23:1176–1180.

Shin DS, Park JS, Lee S-B, Lee S-H, Chung J, Chung MS. 2009b.Surface model of the gastrointestinal tract constructed from theVisible Korean. Clin Anat 22:601–609.

Shin DS, Park JS, Park HS, Hwang SB, Chung MS. 2012d. Outlin-ing of the detailed structures in sectioned images from Visible Ko-rean. Surg Radiol Anat 34:235–247.

Shin DS, Park JS, Shin BS, Chung MS. 2011c. Surface models ofthe male urogenital organs built from the Visible Korean usingpopular software. Anat Cell Biol 44:151–159.

Sorensen MS, Mosegaard J, Trier P. 2009. The visible ear simulator:a public PC application for GPU-accelerated haptic 3D simulationof ear surgery based on the visible ear data. Otol Neurotol30:484–487.

Stimmer H. 2011. Enlargement of the cochlear aqueduct: does itexist? Eur Arch Otorhinolaryngol 268:1655–1661.

Wang H, Merchant SN, Sorensen MS. 2007. A downloadable three-dimensional virtual model of the visible ear. ORL J Otorhinolar-yngol Relat Spec. 69:63–67.

Wang H, Northrop C, Burgess B, Liberman MC, Merchant SN.2006. Three-dimensional virtual model of the human temporalbone: a stand-alone, downloadable teaching tool. Otol Neurotol27:452–457.

1186 PARK ET AL.