Accepted Manuscript

Title: Stereoscopic neuroanatomy lectures using athree-dimensional virtual reality environment

Author: Ralf A. Kockro MD Christina Amaxopoulou MDTim Killeen MRCS Wolfgang Wagner MD Robert Reisch MDAngelika Gutenberg MD Alf Giese MD Eckart Stofft MDAxel T. Stadie MD

PII: S0940-9602(15)00084-9DOI: http://dx.doi.org/doi:10.1016/j.aanat.2015.05.006Reference: AANAT 50963

To appear in:

Received date: 10-9-2014Revised date: 25-5-2015Accepted date: 27-5-2015

Please cite this article as: Kockro, R.A., Amaxopoulou, C., Killeen, T., Wagner, W.,Reisch, R., Gutenberg, A., Giese, A., Stofft, E., Stadie, A.T.,Stereoscopic neuroanatomylectures using a three-dimensional virtual reality environment., Annals of Anatomy(2015), http://dx.doi.org/10.1016/j.aanat.2015.05.006

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 28

Accep

ted

Man

uscr

ipt

Stereoscopic neuroanatomy lectures using a three-dimensional virtual reality environment.

Ralf A Kockro MD,1,3 Christina Amaxopoulou MD,2 Tim Killeen MRCS,1,4 Wolfgang Wagner MD,3 Robert Reisch MD,1,3 Angelika Gutenberg MD,3 Alf Giese MD,3 Eckart Stofft MD,5 Axel T Stadie MD.3

1. Department of Neurosurgery, Klinik Hirslanden, Zurich, Switzerland

2. Department of Radiology, Children’s Hospital Zurich, Switzerland

3. Department of Neurosurgery, University Hospital Mainz, Germany

4. Paraplegia Laboratory, University Hospital Balgrist, Zurich, Switzerland

5. Department of Anatomy, University Hospital, Mainz. Germany

Running title: 3D virtual reality in neuroanatomy teaching

Corresponding author:

Ralf A. Kockro, MDDepartment of NeurosurgeryKlinik HirslandenWitellikerstrasse 408032 ZurichSwitzerlandTel.: +41 44 387 21 30Fax.: +41 44 387 21 31Email: ralf.kockro@hirslanden.ch

Conflict of Interest Statement

Ralf A Kockro is a co-inventor of the Dextroscope Virtual Reality system. He no longer

has any financial interest in the company and has no other conflict of interest to

declare. The other authors declare that they are not aware of any real or potential

conflict of interest in relation to this study.

Page 2 of 28

Accep

ted

Man

uscr

ipt

Abstract

Introduction: Three-dimensional (3D) computer graphics are increasingly used to

supplement the teaching of anatomy. While most systems consist of a program which

produces 3D renderings on a workstation with a standard screen, the Dextrobeam

virtual reality VR environment allows the presentation of spatial neuroanatomical

models to larger groups of students through a stereoscopic projection system.

Materials and Methods: Second-year medical students (n=169) were randomly

allocated to receive a standardised pre-recorded audio lecture detailing the anatomy of

the third ventricle accompanied by either a two-dimensional (2D) PowerPoint

presentation (n=80) or a 3D animated tour of the third ventricle with the DextroBeam.

Students completed a 10-question multiple-choice exam based on the content learned

and a subjective evaluation of the teaching method immediately after the lecture.

Results: Students in the 2D group achieved a mean score of 5.19 (±2.12) compared to

5.45 (±2.16) in the 3D group, with the results in the 3D group statistically non-inferior to

those of the 2D group (p<0.0001). The students rated the 3D method superior to 2D

teaching in four domains (spatial understanding, application in future anatomy classes,

effectiveness, enjoyableness) (p<0.01). Conclusion: Stereoscopically-enhanced 3D

lectures are valid methods of imparting neuroanatomical knowledge and are well

received by students. More research is required to define and develop the role of large-

group VR systems in modern neuroanatomy curricula.

Keywords

Three-dimensional; virtual reality; neuroanatomy; medical education; medical students;

anatomical models; Dextrobeam; third ventricle.

Page 3 of 28

Accep

ted

Man

uscr

ipt

1. Introduction

Internationally, there is a decline in the use of cadaveric dissection as a means

of familiarising medical students with human anatomy, in part driven by the high

cost of maintaining the infrastructure and expertise necessary for dissection

(Gartner, 2003; Korf et al., 2008). There is also evidence that medical schools

are devoting larger proportions of limited curriculum time to, for example, the

development of communication, clinical or research skills, at the expense of

anatomy teaching (Marks, 2002; Sugand et al., 2010; Turney, 2007). These

trends are in turn leading to increased innovation in the way anatomy is taught.

Cadaveric teaching is now supplemented or supplanted by a mixture of

lectures, presentations, physical models and, increasingly, computer

simulations (Tam et al., 2009). While the latter are still dominated by an array of

platforms displaying content in two dimensions (2D), advances in three-

dimensional (3D) imaging have led to the introduction of a number of systems

allowing students to immerse themselves in a virtual reality (VR) environment in

which faithful representations of anatomical regions can be observed and

manipulated.

Most research into the efficacy of such systems and their reception by students

has focused on individual or small group teaching, with 3D models presented on

standard workstations through locally installed programs or using internet-

based, e-learning platforms. Several make use of freely-available datasets,

such as those resulting from the US, Korean and Chinese Visible Human (VH)

projects (Liu et al., 2013; Spitzer et al., 1996), some of which are pre-

Page 4 of 28

Accep

ted

Man

uscr

ipt

segmented (Kockro & Hwang, 2009; Levinson et al., 2007; Liu et al., 2013;

Temkin et al., 2006). In others, reconstructions of clinical or research magnetic

resonance imaging (MRI) or computed tomography scans form the basis of the

models (Gopalakrishnakone et al., 2011; Serra et al., 2002).

While the use of small-group, individual and online teaching using 3D

anatomical models will doubtless increase as the technology and delivery

methods improve, it is likely that medical students will continue to be taught

using a combination of methods, including didactic lectures. No attempt has

hitherto been made to assess the feasibility of presenting 3D models to larger

groups of students to enhance the delivery of such lectures. Although designed

to preoperatively visualise, plan and present neurosurgical interventions in a 3D

environment, the DextroBeam system is a robust and versatile platform for the

display of high resolution, 3D neuroanatomical models via a stereoscopic

projector to large groups (Kockro et al., 2007). We carried out a randomised trial

in which the anatomy of the third ventricle was presented to a large group of

students in a seminar room using a 3D anatomical model or a traditional

PowerPoint presentation with identical content and duration. Efficacy in terms of

anatomical knowledge retained and a subjective evaluation by the students

were recorded.

Page 5 of 28

Accep

ted

Man

uscr

ipt

2. Materials and Methods

Second-year medical students of the University of Mainz, Germany, who had

completed their regular lecture series on cranial anatomy but had yet to begin

their cadaveric dissection neuroanatomy course were recruited to the study.

Participation was voluntary. Students were randomly allocated based on the

initial letter of their surname in groups of approximately 20 to a lecture in which

they would watch either a 2D or 3D presentation. The students participating in

the study did not know the instructors. The sole exclusion criterion was any self-

reported abnormality of stereoscopic vision. Under applicable local legislation,

ethical approval was not required. Permission to carry out the study was

received from the dean’s office.

2.1 2D teaching session

Students allocated to the 2D session watched an automated PowerPoint

presentation (Microsoft, Redmond, USA) in which the topographic anatomy of

the human ventricular system was presented. The presentation was designed

by neurosurgeons with an interest in clinical anatomy teaching. The anatomy

was displayed using images taken from standard textbooks (Fritsch & Kühnel,

2001; Rauber & Kopsch, 1988, 1987; Schünke et al., 2009) and the details

referred to in the commentary were highlighted using arrows and/or coloured

outlines of the respective structures, with the names of the labelled structures

also appearing on the slide. In order to eliminate bias arising from differences in

the speed of delivery of a live, oral presentation, the audio aspect of the

presentation was pre-recorded as an audio file and played in synchrony with the

Page 6 of 28

Accep

ted

Man

uscr

ipt

slides. The script for this audio component (Table 1), in the German language

and consisting of 501 words and created from relevant explanations in three

German anatomy textbooks (Fritsch & Kühnel, 2001; Rauber & Kopsch, 1988,

1987), was recorded using Quicktime (Apple, Cupertino, USA). Playback was

through a laptop connected to the audio system in the seminar room. The

sessions lasted 20 minutes.

2.2 3D teaching session

The 3D teaching session was performed using the DextroBeam system (Bracco

Advanced Medical Technologies, Princeton, USA). This is a stereo-projector

system capable of projecting a stereoscopic image on a special screen. Images

intended for the left and right eye are split using polarising filters placed in front

of the projector’s lens and the 3D image is perceived stereoscopically by

viewers wearing passively polarising glasses. For a detailed discussion of the

system and the technology behind it, please see Serra et al, 2002.

Students were asked to wear stereoscopic glasses for the duration of the

presentation, which was accompanied by the same audio file used in the 2D

presentation. The virtual 3D model (1024 x 768, XGA resolution) used in this

session was created from a cranial MRI and cranial MRI angiogram of a healthy

individual (512 x 512 matrix with 1.5mm slice thickness). The DextroBeam 3D

image processing and segmentation tools were used to create a virtual 3D

ventricular system including all the anatomical structures presented in the 2D

presentation. Anatomical structures mentioned in the audio commentary were

highlighted with the identical colours used for the corresponding structure in the

2D presentation. Using the DextroBeam’s stereoscopic movie recording feature,

Page 7 of 28

Accep

ted

Man

uscr

ipt

we created an animated tour around and through the ventricular anatomy

featuring the same structures, key perspectives and labelling identical to that of

the 2D presentation.

2.3 Multiple-choice questions

Immediately following their teaching session, each participant was asked to

complete a short examination consisting of ten multiple-choice questions

(MCQs) relating to the topographical anatomy of the third ventricle (Table 2).

These questions were not formally validated but were developed and agreed on

by an expert committee of four neurosurgeons and anatomists as being

representative of the level of neuroanatomical knowledge required for success

in the course examinations. They were informally tested on an independent

group of students who had completed their neuroanatomy curriculum.

Participants were given 10 minutes to anonymously answer the questions by

indicating the correct answer. The students then completed an anonymous

questionnaire asking for their age and gender and four questions in which they

could subjectively rate their learning experience (Table 3).

2.4 Statistical methods

Each correct answer in the MCQ examination was awarded one point, with a

maximum of 10 points achievable. There was no negative marking. Statistical

analysis was performed using SPSS v.15 (IBM, Armonk, NY, USA) and JMP

(SAS, Cary, NC, USA). For the analysis of demographic data and the rating

responses, χ-square and a two-sided Fisher’s exact test were employed. A two-

sample, one-sided t-test was used to assess for non-inferiority of the

Page 8 of 28

Accep

ted

Man

uscr

ipt

experimental teaching method. Non-inferiority testing differs from a standard

comparison of means in that the null hypothesis states that the method

assessment (3D teaching) is neither equivalent nor superior to the control

sample (2D teaching). Rejection of the null hypothesis thus confirms the method

being tested is at least equivalent to the sample compared, with equivalence

bounded by a practical equivalence term δ, set a priori in this case as a

difference of 1 point on the MCQ test. Direct comparison of the means was

precluded in our sample as an analysis demonstrated that the sample size

available did not generate sufficient statistical power for an assessment. For all

tests, statistical significance was defined by a p value of ≤ 0.05.

3. Results

3.1 Demographic data

One hundred and sixty-nine students volunteered to take part in this study. As

no student declared an abnormality of stereoscopic vision, all volunteers were

randomised to the 3D or 2D teaching. Eighty-nine (52.7%) watched the 3D, VR

training, with 80 (47.3%) attending the 2D, PowerPoint-based session.

Demographic details were available for all students. Gender distribution was

comparable to that of other German medical schools (Liebhardt et al., 2010),

with 106 (62.7%) female participants. The mean age of students was 22.5 years

(range 20 – 32 years). Students had completed between two and six semesters

of medical school, with a majority (84.6%) in their fourth semester.

3.2 MCQ examination results

The mean score of the ten-question MCQ test performed by students who

watched the 2D presentation was 5.19 (median 5; S.D. 2.12). In the group

Page 9 of 28

Accep

ted

Man

uscr

ipt

receiving 3D teaching, the mean score was 5.45 (median 6; S.D. 2.16). The 3D

presentation was shown to be statistically non-inferior to traditional 2D

presentation (p<0.0001). Comparison of the two means showed a non-

significant trend (p=0.215) towards superiority of 3D teaching, although the

study was underpowered to definitively detect any difference, so this result

should be interpreted with caution. Non-parametric subgroup analysis based on

gender (p=0.145), age group (p=0.458) and number of completed semesters

(p=0.904) did not reveal any significant difference between the groups.

3.3 Results of evaluation questionnaire

The evaluation questionnaire was completed by 100% of student participants.

This high rate of compliance was achieved by presenting the students with the

questionnaire immediately after the MCQ test. Students consistently and

significantly rated the 3D VR presentation sessions more highly compared to

usual teaching than the 2D presentation across all four evaluation domains

(Figure 2; Fisher’s exact test; p<0.001 in all cases).

4. Discussion

Learning neuroanatomy involves the mastery of 3D spatial information as

applied to the structures and relations of the human nervous system. Acquiring

such spatial knowledge necessarily requires the initial perception of 3D

information from which a 3D mental image is constructed by each learner and

which may be continually augmented and refined through further learning and

validation (Marks, 2000). While 3D information may be synthesised from 2D

sources such as books and traditional lectures, one of the strengths of

Page 10 of 28

Accep

ted

Man

uscr

ipt

cadaveric dissection is that spatial relations and complex structures can be

viewed and manipulated in context (Marks, 2000; Older, 2004).

Several groups have presented and/or evaluated VR teaching systems for

undergraduate education. In 2006, Temkin et al described an interactive web-

based 3D model based on the VH dataset, some elements of which were

rendered in blue-red stereoscopic 3D (Temkin et al., 2006). In the same year,

the VH dataset was also used to create an interactive, fully 3D model of the

human temporal bone which was used for neurosurgical training (Kockro &

Hwang 2009). Neither of these systems were subjected to randomised

evaluation. Subsequent trials of related systems yielded mixed results, with VR

teaching of the carpal bones (Garg et al., 1999; Garg et al. 2002; Garg et al.,

2001) and shoulder (Hariri, et al., 2004) not convincingly assessed as superior

to standard teaching, while similar studies in which the middle ear (Nicholson et

al., 2006; Venail et al., 2010) and biliary system (Beermann et al., 2010) were

taught showed a clear benefit of VR. In such small-group settings, VR and other

computer-based methods may not be superior to anatomical models (Khot et

al., 2013; Preece et al., 2013), which are becoming increasingly diverse,

detailed and lifelike as costs fall and 3D printing methods open up new

possibilities.

Such physical models and their virtual counterparts, viewed using workstation-

based or online 3D systems, seek to reproduce the three-dimensionality of

dissection and are now established educational adjuncts in many medical

schools. One disadvantage of such systems, however, is their limited scalability,

with the maximum number of contemporaneous participants limited to around

Page 11 of 28

Accep

ted

Man

uscr

ipt

five. Another drawback, perhaps paradoxically, is that it seems that the more

freedom students are allowed in manipulating and exploring models on their

own, the higher the risk that the learning experience is degraded by time spent

off-task or through key concepts being overlooked (Garg et al., 1999; Levinson

et al., 2007; Phelps et al., 2004; Silén et al., 2008). Many of these tools are

based on concepts allied to the discovery- and enquiry-based schools of

learning theory, which hold that students should be given the tools and freedom

to gain knowledge themselves (Barrows, 1986). Critiques of this approach

argue that expert guidance during the learning process is necessary,

particularly when knowledge is first being accumulated, as students lack the

basic framework to “internally” guide their own learning (Kirschner et al., 2006).

Delivering 3D neuroanatomical content to a larger group as a stereoscopically-

enhanced animated lecture, as described here, allows students to visualise and

lay down memories of structures in 3D while receiving instruction from an expert

neuroanatomist who can ensure the pertinence of the delivered content. While

the presentations used in the study were fully standardised in order to reduce

bias, in everyday use, 3D lectures need not lack interactivity – lecturers can

respond to questions or requests for clarification and illustrate their responses

by manipulating the model.

Despite the relative simplicity of the 3D computer models used in this study –

lacking shadows, texture, detail and haptic elements – the results of the MCQ

test demonstrated that VR teaching in neuroanatomy is at least equivalent in

terms of efficacy to PowerPoint presentations based on elaborate, yet 2D,

textbook representations. The stereoscopic three-dimensionality of VR teaching

Page 12 of 28

Accep

ted

Man

uscr

ipt

directly incorporates the intrinsic spatial relationships of the anatomical sites

studied, and thus may confer a spatial knowledge advantage which is beyond

the ability of MCQs to reliably test. Habitual exposure to this critical aspect of

neuroanatomy during undergraduate training may catalyse the acquisition of the

practical anatomical awareness required, for instance, in a postgraduate

surgical career.

Students exposed to VR teaching in this study overwhelmingly approved of its

use. Three-quarters of students rated the ability of the 3D lecture to convey

spatial understanding as good or very good. This is significant, as traditional

didactic lectures are rated poorly on this aspect, with this method coming last in

a survey of six different anatomy teaching techniques (Kerby et al., 2011), a

finding replicated in the 2D group in this study (24% good / very good). While

the novelty of the 3D VR teaching experience against a possible background of

“PowerPoint fatigue” – presentations using the Microsoft software are

ubiquitous in medical schools – may explain some of the enthusiasm, this

striking finding is in keeping with other studies evaluating the effect of individual

/ small-group VR teaching on student approval and motivation (Beermann et al.,

2010; Garg et al., 2002; Murgitroyd et al., 2015; Silén et al., 2008; Venail et al.,

2010; for a meta-analysis, see Yammine & Violato, 2014). Students motivated

and engaged by innovative learning methods are more likely to learn and retain

knowledge (Battulga et al., 2012; Huang et al., 2010; Shim et al., 2003) and

more engaging teaching may improve learning outcomes indirectly, through

increased overall motivation, improved attendance and inspiring students to use

novel methods in their self-directed learning.

Page 13 of 28

Accep

ted

Man

uscr

ipt

The cost of stereoscopic projectors and polarising glasses have dropped

significantly in recent years and the main cost of acquiring and using such

systems now arises from the proprietary software and expertise necessary to

construct the models. The DextroBeam system is only a step in the rapid

evolution of VR teaching platforms, with novel systems enabling the display of

stereoscopic models without the need for special glasses in development

(Abildgaard et al., 2010). We are currently developing a further teaching

intervention with an upgraded Dextrobeam setup and publication of this data will

hopefully include accompanying videos.

The potential advantages of VR teaching are not limited to undergraduate

application. With limitations on junior surgeons’ working hours and fewer

opportunities to learn surgical anatomy in the operating theatre, immersive 3D

anatomy learning has a role to play in supplementing more traditional methods

at all levels of training (Coulter & Brennan, 2013; Kirkman et al., 2014; Kockro,

2013; Lemole et al., 2007; Murgitroyd et al., 2015; Schirmer et al., 2013)

Previous studies assessing small-group 3D VR anatomy have shown that

individual spatial ability is a predictor of improved learning outcomes following

3D learning (Berney et al., 2015; Phelps et al., 2004; Yammine & Violato, 2014)

and by extension it has been suggested that male students may benefit more

from 3D teaching than females (Battulga et al., 2012; Beermann et al., 2010).

While we did not assess our participants’ spatial ability, our limited subgroup

analysis did not show any such gender effect, nor did age or number of medical

school semesters completed show an association with MCQ results.

Page 14 of 28

Accep

ted

Man

uscr

ipt

Several factors not tested in this study may account for the apparent

discrepancy between the students’ subjectively excellent learning experience

and the statistically indiscernible MCQ results. The study only had one

assessment time-point, immediately after teaching. It may be that the 3D

context provided in stereoscopically-enhanced lectures allows for better

retention of anatomical knowledge over time, meaning the benefit of 3D over 2D

lectures would become apparent only later. Proof of this hypothesis may reveal

a key advantage of 3D teaching and could be achieved by reassessing students

at multiple time-points, something that was beyond the scope of the current

study.

Furthermore, students were not rewarded for good performance in the MCQ

examination, nor did it count towards their grade or chances of progression.

Participation was presented as an opportunity for formative assessment on an

area which was to be tested at the end of the semester. As such, the students

may not have been fully motivated to apply themselves in the test and this may

account for the relatively low mean scores (52-55%) in both groups, although

the test was relatively demanding and consisted of several ranking and

matching questions which must be answered perfectly to score a point.

Finally, the anatomy of the third ventricle was thought to lend itself to 3D

teaching due to its moderate complexity, difficulty to visualise and relevance to

both undergraduate neuroanatomy and neurosurgery (Schirmer et al., 2013).

Previous studies in which superiority of 3D teaching was demonstrated tended

to use more complex regions, such as the temporal bone and biliary tree, as

Page 15 of 28

Accep

ted

Man

uscr

ipt

their substrate (Beermann et al., 2010; Nicholson et al., 2006), while negative

findings were seen using simpler structures. Future studies of 3D teaching

should explore this effect by comparing regions of different complexity, while

also assessing the role of the dimensionality of the teaching substrate on

knowledge retention over serial time-points. The optimal role of VR teaching

may ultimately lie in offering expert teachers the ability to not only demonstrate

the anatomy of a region, but also how to approach and manipulate its structures

efficiently and safely in 3D, something very difficult to achieve with static 2D

images or animations.

5. Conclusion

This study demonstrates that the use of stereoscopic, computer-generated,

anatomical models delivered using a 3D projection system can enhance

undergraduate neuroanatomy lectures. The 3D presentation, despite being

based on a computer model simplified to the extent of merely illustrating the

core 3D anatomical relationships, was shown to be at least as effective as a 2D

presentation using classical textbook drawings. These findings, combined with

the clear preference of the students in favour of the 3D teaching method, should

be taken as encouragement to continue to develop more elaborate, realistic

virtual models and 3D courseware with the aim of providing a high-fidelity,

virtual teaching environment transmitting anatomical knowledge in its inherent,

3D form.

Acknowledgements

Page 16 of 28

Accep

ted

Man

uscr

ipt

The authors thank the medical students of Mainz University who agreed to take

part in this study for their time and enthusiasm. The Clinical Research Priority

Program (CRPP) for Neuro-Rehab of the University of Zurich provided

resources for Tim Killeen.

References

Abildgaard, A., Witwit, A.K., Karlsen, J.S., Jacobsen, E.A., Tennøe, B., Ringstad, G., Due-Tønnessen, P., 2010. An autostereoscopic 3D display can improve visualization of 3D models from intracranial MR angiography. Int. J. Comput. Assist. Radiol. Surg. 5, 549–54.

Barrows, H.S., 1986. A taxonomy of problem-based learning methods. Med. Educ. 20, 481–486.

Battulga, B., Konishi, T., Tamura, Y., Moriguchi, H., 2012. The effectiveness of an interactive 3-dimensional computer graphics model for medical education. Interact. J. Med. Res. 1, e2.

Beermann, J., Tetzlaff, R., Bruckner, T., Schöebinger, M., Müller-Stich, B.P., Gutt, C.N., Meinzer, H.-P., Kadmon, M., Fischer, L., 2010. Three-dimensional visualisation improves understanding of surgical liver anatomy. Med. Educ. 44, 936–40.

Berney, S., Bétrancourt, M., Molinari, G., Hoyek, N., 2015. How spatial abilities and dynamic visualizations interplay when learning functional anatomy with 3D anatomical models. Anat. Sci. Educ. Feb 16, 1-7.

Coulter, I., Brennan, P., 2013. Simulation in Neurosurgery: A Survey of Experiences and Perceptions in the UK. Bull. R. Coll. Surg. Engl. 95, 304–307.

Fritsch, H., Kühnel, W., 2001. Taschenatlas der Anatomie, Band 3, Nervensystem und Sinnesorgane, 7th Editio. ed. Thieme, Stuttgart.

Garg, A., Norman, G., Spero, L., Maheshwari, P., 1999. Do virtual computer models hinder anatomy learning? Acad. Med. 74, S87-9.

Garg, A., Norman, G., Sperotable, L., 2001. How medical students learn spatial anatomy. Lancet 357, 363–364.

Garg, A.X., Norman, G.R., Eva, K.W., Spero, L., Sharan, S., 2002. Is there any real virtue of virtual reality?: the minor role of multiple orientations in learning anatomy from computers. Acad. Med. 77, S97–9.

Page 17 of 28

Accep

ted

Man

uscr

ipt

Gartner, L.P., 2003. Anatomical sciences in the allopathic medical school curriculum in the United States between 1967-2001. Clin. Anat. 16, 434–9.

Gopalakrishnakone, P., Jianfeng, L., Sun, G.P., Abeykoon, A., Noel, O., Fernando, N., Cheok, A.D., 2011. A multimodal virtual anatomy learning tool for medical education. [Online] Available from http://www.academia.edu/1268456/A_Multimodal_Virtual_Anatomy_Learning_Tool_for_Medical_Education. Accessed 17th April 2015.

Hariri, S., Rawn, C., Srivastava, S., Youngblood, P., Ladd, a, 2004. Evaluation of a surgical simulator for learning clinical anatomy. Med. Educ. 38, 896–902.

Huang, H.-M., Rauch, U., Liaw, S.-S., 2010. Investigating learners’ attitudes toward virtual reality learning environments: Based on a constructivist approach. Comput. Educ. 55, 1171–1182.

Kerby, J., Shukur, Z.N., Shalhoub, J., 2011. The relationships between learning outcomes and methods of teaching anatomy as perceived by medical students. Clin. Anat. 24, 489–97.

Khot, Z., Quinlan, K., Norman, G.R., Wainman, B., 2013. The relative effectiveness of computer-based and traditional resources for education in anatomy. Anat. Sci. Educ. 6, 211–215.

Kirkman, M. a, Ahmed, M., Albert, A.F., Wilson, M.H., Nandi, D., Sevdalis, N., 2014. The use of simulation in neurosurgical education and training. J. Neurosurg. 121, 228–46.

Kirschner, P.A., Sweller, J., Clark, R.E., 2006. Why minimal guidance during instruction does not work: an analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educ. Psychol. 41, 75–86.

Kockro, R., Hwang, P., 2009. Virtual Temporal Bone: An Interactive

3‐Dimensional Learning Aid for Cranial Base Surgery. Neurosurgery 64,

216–230.

Kockro, R.A., 2013. Neurosurgery Simulators - Beyond the Experiment. World Neurosurg. 80, 120–121.

Kockro, R.A., Stadie, A., Schwandt, E., Reisch, R., Charalampaki, C., Ng, I., Yeo, T.T., Hwang, P., Serra, L., Perneczky, A., 2007. A collaborative virtual reality environment for neurosurgical planning and training. Neurosurgery 61, 379–91.

Page 18 of 28

Accep

ted

Man

uscr

ipt

Korf, H.-W., Wicht, H., Snipes, R.L., Timmermans, J.-P., Paulsen, F., Rune, G., Baumgart-Vogt, E., 2008. The dissection course - necessary and indispensable for teaching anatomy to medical students. Ann. Anat. 190, 16–22.

Lemole, G.M., Banerjee, P.P., Luciano, C., Neckrysh, S., Charbel, F.T., 2007. Virtual reality in neurosurgical education: part-task ventriculostomy simulation with dynamic visual and haptic feedback. Neurosurgery 61, 142–8.

Levinson, A.J., Weaver, B., Garside, S., McGinn, H., Norman, G.R., 2007. Virtual reality and brain anatomy: a randomised trial of e-learning instructional designs. Med. Educ. 41, 495–501.

Liebhardt, H., Fegert, J., Dittrich, W., Nürnberger, F., 2010. Medizin studieren mit Kind. Ein Trend der Zukunft. Dtsch Arztebl 107, 1613–1615.

Liu, K., Fang, B., Wu, Y., Li, Y., Jin, J., Tan, L., Zhang, S., 2013. Anatomical education and surgical simulation based on the Chinese Visible Human: a three-dimensional virtual model of the larynx region. Anat. Sci. Int. 88, 254–258.

Marks, S.C., 2002. Clinical Anatomy for Health Care and Medical Education in the 21 st Century. J. Japanese Res. Soc. Clin. Anat. 1, 4–7.

Marks, S.C., 2000. The role of three-dimensional information in health care and medical education: the implications for anatomy and dissection. Clin. Anat. 13, 448–52.

Murgitroyd, E., Marduska, M., Gonzalez, J., Watson, A., 2015. 3D digital anatomy modelling – Practical or pretty? Surgeon. 2–5.

Nicholson, D.T., Chalk, C., Funnell, W.R.J., Daniel, S.J., 2006. Can virtual reality improve anatomy education? A randomised controlled study of a computer-generated three-dimensional anatomical ear model. Med. Educ. 40, 1081–7.

Older, J., 2004. Anatomy: A must for teaching the next generation. Surgeon. 2, 79–90.

Phelps, A., Fritchle, A., Hoffman, H., 2004. Passive vs. active virtual reality learning: The effects on short- and long-term memory of anatomical structures. Stud. Health Technol. Inform. 98, 298-300.

Preece, D., Williams, S.B., Lam, R., Weller, R., 2013. “Let’s Get Physical”: Advantages of a physical model over 3D computer models and textbooks in learning imaging anatomy. Anat. Sci. Educ. 6, 216–224.

Rauber, A., Kopsch, F., 1988. Anatomie des Menschen Lehrbuch und Atlas, Band IV: Topographie der Organsysteme, Systematik der Leitungsbahnen. Thieme, Stuttgart.

Page 19 of 28

Accep

ted

Man

uscr

ipt

Rauber, A., Kopsch, F., 1987. Anatomie des Menschen Lehrbuch und Atlas, Band III. Nervensystem Sinnesorgane. Thieme, Stuttgart.

Schirmer, C.M., Elder, J.B., Roitberg, B., Lobel, D. a., 2013. Virtual reality-based simulation training for ventriculostomy: An Evidence-based approach. Neurosurgery 73, 66–73.

Schünke, M., Schulte, E., Schumacher, U., Voll, M., Wesker, K., 2009. Prometheus LernAtlas der Anatomie: Kopf, Hals und Neuroanatomie, 2nd Editio. ed. Thieme, Stuttgart.

Serra, L., Kockro, R., Goh, L.C., Ng, H., Lee, E.C.K., 2002. The DextroBeam: a stereoscopic presentation system for volumetric medical data. Stud. Health Technol. Inform. 85, 478–84.

Shim, K.-C., Park, J.-S., Kim, H.-S., Kim, J.-H., Park, Y.-C., Ryu, H.-I., 2003. Application of virtual reality technology in biology education. J. Biol. Educ. 37, 71–74.

Silén, C., Wirell, S., Kvist, J., Nylander, E., Smedby, O., 2008. Advanced 3D visualization in student-centred medical education. Med. Teach. 30, e115–e124.

Spitzer, V., Ackerman, M.J., Scherzinger, a L., Whitlock, D., 1996. The visible human male: a technical report. J. Am. Med. Informatics Assoc. 3, 118–130.

Sugand, K., Abrahams, P., Khurana, A., 2010. The anatomy of anatomy: a review for its modernization. Anat. Sci. Educ. 3, 83–93.

Tam, M., Hart, A., Williams, S., 2009. Is learning anatomy facilitated by computer-aided learning? A review of the literature. Med. Teach. 31, 393–396.

Temkin, B., Acosta, E., Malvankar, A., Vaidyanath, S., 2006. An interactive three-dimensional virtual body structures system for anatomical training over the internet. Clin. Anat. 19, 267–74.

Turney, B.W., 2007. Anatomy in a modern medical curriculum. Ann. R. Coll. Surg. Engl. 89, 104–7.

Venail, F., Deveze, A., Lallemant, B., Guevara, N., Mondain, M., 2010. Enhancement of temporal bone anatomy learning with computer 3D rendered imaging software. Med. Teach. 32, e282–8.

Yammine, K., Violato, C., 2014. A meta-analysis of the educational effectiveness of three-dimensional visualization technologies in teaching anatomy. Anat. Sci. Educ. Dec 31, 1-14.

Page 20 of 28

Accep

ted

Man

uscr

iptTable 1. Full transcript of accompanying audio track played alongside both

2D and 3D presentations. Translated from the original German.

Along its lateral axis, the unpaired third ventricle is small, but is relatively extensive in the dorso-ventral plane. The ventricle has a floor and, from the roof posteriorly, expansions known as recesses. Rostrally, the third ventricle is connected to both lateral ventricles through the intraventricular foraminae. Caudally, the ventricle is continuous with the mesencephalic cerebral aqueduct although the longitudinal axis of the ventricle – a line connecting the rostral and posterior epithalamic commisures – is angulated ventrally with respect to the long axis of the aqueduct.

The roof of the third ventricle consists of the tela chorioidea with the choroid plexus of the third ventricle. The latter continues as the choroid plexus of the lateral ventricle in the region of the intraventricular foramina. In the epithalamus, the suprapineal recess arches dorsally over the pineal gland. The pineal recess forms a shallow hollow in the pineal gland. Above the pineal recess cross the fibres of the habenular commissure, while the epithalmic commissure lies under the recess.

The anterior wall of the third ventricle is formed by the lamina terminalis. The lamina terminalis is bounded caudally by the optic recess, formed from the horizontal bulge produced by the optic chiasm. Behind the optic chiasm, the funnel-shaped infundibulum and the infundibular recess form a convex impression in the floor of the ventricle.

The floor of the ventricle is part of the hypothalamus and consists mainly of a thin, grey lamella. The infundibulum continues as the pituitary stalk. Behind the infundibulum lies the tuber cinereum, beyond which are found the paired mammillary bodies. In the large, upper part of the lateral walls, the thalamus bulges out in the form of a flat oval. In 75% of cases, at variable positions and to varying degrees, the two thalami are connected through the ventricular lumen by the interthalamic adhesion (massa intermedia). Caudal to the floor of the third ventricle, in the angle formed by the floor of the midbrain, the cerebral peduncles and the pons, lies the interpeduncular cistern before which, surrounding the chiasm, is the chiasmatic cistern. In the interpeduncular cistern the basilar artery bifurcates in the interpendicular fossa behind the mammillary bodies into the two posterior cerebral arteries. In relation to the skull base, this bifurcation is usually at the level of the dorsum sellae.

Page 21 of 28

Accep

ted

Man

uscr

ipt

Table 2. Multiple choice questions on the applied anatomy of the third

ventricle. Correct responses in bold italic type.

1. Order the following epithalamic structures in the correct order, from cranial to caudal.

1. Cerebral aqueduct2. Suprapineal recess3. Habenular commissure4. Pineal recess

a. 1-3-4-2b. 2-3-1-4c. 4-1-3-2d. 2-3-4-1e. 3-4-1-2

2. Which structure lies beneath the floor of the third ventricle and may be injured during its perforation?

a. Lamina terminalisb. Basilar arteryc. Interventricular foramen of

Monroed. Optic chiasme. Thalamus

3. Which structure often hinders direct access from the interventricular foramen to the cerebral aqueduct?

a. Interthalamic adhesionb. Thalamusc. Commisura habenularumd. Pineal bodye. Optic chiasm

4. Which response(s) is/are correct?1. The neurosurgeon can gain access to the

third ventricle by opening the lamina terminalis.

2. Access from the fourth ventricle to the third ventricle can be gained through the interventricular foramen.

3. The thalamus can be found in the floor of the third ventricle.

a. 1 and 3 are correctb. 1, 2 and 3 are correctc. Only 1 is correctd. Only 2 is correcte. No response is correct

5. Which structure can the surgeon follow through the interventricular foramen from the lateral to the third ventricle?

a. Pineal recessb. Choroid plexusc. Epithalamic commissured. Infundibulume. Mammillary bodies

6. Which structure can be found in the floor of the third ventricle?

1. Infundibular recess2. Interthalamic adhesion3. Pineal recess4. Mammillary bodies

a. 1, 3 and 4 are correctb. 2 and 4 are correctc. Only 4 is correctd. Only 2 is correcte. 1 and 4 are correct

7. Which structure is found along the caudal extension of a line drawn between the interventricular foramen and the tubercinereum?

a. Choroid plexusb. Suprapineal recessc. Lamina terminalisd. Interpeduncular cisterne. Cerebral aqueduct

Page 22 of 28

Accep

ted

Man

uscr

ipt

8. Order these anatomical structures in the following order: cranial – caudal – rostral –dorsal – lateral

1. Lamina terminalis2. Mammillary bodies3. Thalamus4. Choroid plexus5. Pineal recess

a. 1-2-3-4-5b. 4-2-1-5-3c. 4-2-5-1-3d. 4-5-3-2-1e. 2-4-3-5-1

9. Which of the following statements is correct?

1. The epithalamic commissure forms the upper border of the entrance of the cerebral aqueduct.

2. The habenular commissure merges with the pineal body.

3. The pineal body lies dorsal to the pineal recess.

a. All statements are correctb. Only 1 is correctc. Only 2 is correctd. Only 3 is correcte. Statements 1 and 3 are

correct

10. The choroid plexus is found on which aspect of the third ventricle?

a. Cranialb. Caudalc. Laterald. Anteriore. Posterior

Table 3. Subjective evaluation questions given completed by all students

immediately after either 2D or 3D neuroanatomy teaching.

Question Response 2D group; n(%) 3D group; n(%)1. This teaching method’s usefulness in imparting spatial understanding of the topographical anatomy of the third ventricle is…

Very goodGoodAveragePoorVery Poor

7 (8.8)12 (15)48 (60)9 (11.3)4 (5)

15 (16.9)52 (58.4)21 (23.6)0 (0)0 (0)

2. This method of teaching should be used in future courses.

Agree stronglyAgree somewhatNeutralDisagree somewhatDisagree strongly

18 (22.5)21 (26.3)15 (18.8)24 (30)2 (2.5)

58 (65.2)28 (31.5)2 (2.2)0 (0)0 (0)

3. This method of teaching anatomy is an effective and practical tool for teaching anatomy.

Agree stronglyAgree somewhatNeutralDisagree somewhatDisagree strongly

23 (28.7)20 (25)8 (10)28 (35)1 (1.3)

70 (78.7)19 (21.3)0 (0)0 (0)0 (0)

4. This method of teaching was enjoyable.

Agree stronglyAgree somewhatNeutralDisagree somewhatDisagree strongly

12 (15)12 (15)33 (41.3)21 (26.3)2 (2.5)

19 (21.3)50 (56.2)18 (20.2)2 (2.2)0 (0)

Page 23 of 28

Accep

ted

Man

uscr

iptFigure Legends

Figure 1. Screenshots from the 2D presentation, delivered using Microsoft

PowerPoint. The presentation was played automatically. A - D show examples

of the slides used in order of presentation. Structures were labelled, with the

wording used consistent between 2D and 3D presentations.

Figure 2. Screenshots from the 3D presentation, delivered using the

DextroBeam system. A) 3D overview of the position of the third ventricle within

the head. The automatic presentation then proceeds to zoom to the region of

interest (B), before proceeding to present the anatomical structures in turn (Fig.

3). Colouring is as follows: off-white; skull base, light blue; ventricular system,

yellow; optic nerves, red; blood vessels, green (anterior); tuber cinereum, green

(posterior); pineal body, dark blue; recessus opticus.

Figure 3. Further screenshots from the 3D presentation. The presentation

moves smoothly between key views in a “tour” of the anatomy consistent in

content and order with those of the 2D presentation. Colouring is as follows: off-

white; skull base, yellow; cranial nerves, red; blood vessels, green (anterior);

tuber cinereum, green (posterior); pineal body, blue; recessus opticus, lilac;

pituitary stalk, purple; choroid plexus, pink; mammillary bodies. As the

screenshots are from a 3D tour, in some images (notably Fig. 3, D), the targets

of the labelling are partially obscured by overlapping imagery. When viewed in

Page 24 of 28

Accep

ted

Man

uscr

ipt

stereoscopic 3D, the labels are aligned with their respective targets so that

clarity is maintained.

Figure 4. Results of the MCQ test showing the percentage of students in

each group achieving particular scores out of 10.

Page 25 of 28

Accep

ted

Man

uscr

ipt

Figure 1

Page 26 of 28

Accep

ted

Man

uscr

ipt

Figure 2

Page 27 of 28

Accep

ted

Man

uscr

ipt

Figure 3

Page 28 of 28

Accep

ted

Man

uscr

ipt

Figure 4