274 SURGERY

THE LEARNING AND IDENTIFICATION OF ANATOMY is acentral component in surgical education and tra-ditionally, anatomic nomenclature is distinct ateach stage of medical training. Excluding basicundergraduate course work, the surgeon’s first for-mal exposure to hepatic nomenclature will mostlikely begin as a first-year medical student. At somelater point, surgical residents will be exposed toCouinaud hepatic segmental anatomy and surgical(resection) hepatic anatomy.

In a necessary move to standardize the termi-nology of liver anatomy and liver resections, theInternational Hepato-Pancreato-Biliary Associat-ion (IHPBA) convened a committee with the resul-tant accepted terminology supported by theIHPBA.1 Although this can be seen as positiveprogress, for the junior and midlevel surgical resi-

dent who will likely not be able to participate in ahepatic case until close to or at chief resident level,hepatic nomenclature remains an imaginary pur-suit in a nontransparent organ. The following edu-cational challenges are not unique to hepaticoperations: limited availability of biological mate-rials, limited availability of expert educators,increasingly specialized procedures, and rapidexpansion of knowledge.

Recognizing these difficulties and experiencingfirsthand the specific challenges in teaching surgi-cal residents how to appropriately interpret hepat-ic radiographic images for operative planning ledus to apply state-of-the-art technology to the prob-lem. Our purpose was to share essential anatomicknowledge for assessment and preoperative plan-ning with multiple residents simultaneously atmultiple physical locations by providing: (1) inter-action with a single master surgeon; (2) visual def-initions of anatomic nomenclature; and (3) avisual understanding of the spatial relationshipsbetween the intrahepatic biliary ducts and portaland hepatic veins.

METHODSMultiple commercial and public domain soft-

ware libraries were used to program our teleim-mersive applications. We use high-performance

Virtual reality: Immersive hepatic surgery educational environmentJonathan C. Silverstein, MD, FACS, Fred Dech, MFA, Marcia Edison, PhD, Peter Jurek, MAMS, W. Scott Helton, MD, FACS, and N. Joseph Espat, MD, FACS, Chicago, Ill

Background. Understanding the spatial relationships among the liver segments, and intrahepatic portaland hepatic veins is essential for surgical treatment of liver diseases. Teleimmersive virtual realityenables improved visualization over conventional media because it supports stereo vision, viewer-centeredperspective, large angles of view, and interactivity with remote locations. We report a successful pilotstudy teaching hepatic surgical principles using teleimmersion.Methods. We developed a teleimmersive environment for teaching with biomedical models including virtual models of the liver segments and portal and hepatic veins. Using the environment, 1 instructorgave a workshop to 6 senior general surgery residents at 2 physical locations. A 24-question (36-point)examination was administered before and after the workshop.Results. The workshop produced significant improvements in the mean test scores between the pretestsand posttests (17.67 to 23.67, P < .02). We found no differences between residents who were with theinstructor and those at the remote location. Six-month delayed testing demonstrated complete retention of new knowledge.Conclusions. The teleimmersive environment enabled surgeons to overcome some of the barriers to teach-ing complex surgical anatomic principles. Using teleimmersive environments, surgical educators andtrainees can interact from locations worldwide using virtual anatomic information to achieve their edu-cational goals. (Surgery 2002;132:274-7.)

From the Departments of Surgery, University of Illinois at Chicago and the University at Chicago, Chicago, Ill

Supported by federal funds from the National Library ofMedicine, National Institutes of Health, under Contract No.N01-LM-9-3543 and under Grant No. R01-LM-06756-01.

Presented at the 63rd Annual Meeting of the Society ofUniversity Surgeons, Honolulu, Hawaii, February 14-16, 2002.

Reprint requests: Jonathan C. Silverstein, MD, FACS, Center forClinical Information, The University of Chicago, Room A-105,MC 6051, 5841 S Maryland Ave, Chicago, IL 60637.

© 2002, Mosby, Inc. All rights reserved.

0039-6060/2002/$35.00 + 0 11/6/125723

doi:10.1067/msy.2002.125723

Surgery Silverstein et al 275Volume 132, Number 2

networks (minimum 100 megabit per second), SGIOnyx2 graphics supercomputers (Mountain View,Calif), and ImmersaDesk interactive virtual realitydisplays (Fakespace Systems, Kitchener, Ontario,Canada).2 Our hardware cost over $250,000, but anenvironment can currently be built that will sup-port our applications for under half our cost, andthese costs continue to decrease. Using these tech-nologies enables 3-dimensional (3D) visualizationof hepatic surface landmarks and intrahepaticstructures with: (1) stereovision, (2) larger-than-lifesize (comfortable viewing angles), (3) viewer-cen-tered perspective (minimizing distortion by dis-playing correctly for each station through the useof head-tracking devices), and (4) remote userinteractivity (smooth frame rates, immediateresponsiveness, and streaming audio for bothteacher and remote students).

Several image processing and illustration soft-ware programs, in coordination with expert surgi-cal knowledge, were used to generate accurate livermodels from the Visible Human Project female.3

We have previously reported our development of asimilar immersive pelvic floor model4 and a fullyfeatured teleimmersive application for interactive,collaborative investigation of arbitrary radiologicaldatasets.5

The immersive hepatic surgery educational envi-ronment (IHSEE) uses several of our most usefulfeatures enabled for both teacher and student con-trol (Fig 1). The liver can be rotated in unlimitedfashion in all directions, moved to any location inthe large virtual space, and increased and decreasedin size using natural “click-and-drag” movement of a6-degrees-of-freedom wand (3D mouse). Thesemovements highlight for the viewer the important“true” anatomic location of hepatic surgical land-marks that are useful for planning and performinghepatic operation (ie, with the liver cranially rotated30 degrees beyond the anatomic position, theFissure of Ganz, gallbladder fossa, and inferior venacava can be demonstrated, underscoring to the view-er the importance of adequate hepatic mobilizationand the relational proximity of major vascular struc-

Fig 1. Upper left shows all models with transparent parenchyma. Upper right shows user deleting segment 4parenchyma models. Lower left shows second user’s wand avatar appearing and rotating models. Lower rightshows second user pointing out right hepatic vein being deleted by first user.

276 Silverstein et al SurgeryAugust 2002

tures). The liver is anatomically divided intoCouinaud segments, with a viewer-enabled ability toadd and delete individual segment parenchyma,portal structures, and hepatic veins. This featureenables an appreciation of the “why” for the depen-dence between segments (segments VI and VIIbeing the right posterior section or segments, V andVIII being the right anterior section). Deletion of asegment or restoration of a segment enables thevisualization of the underlying biliary and portal andhepatic venous anatomy. In any position, any size,and at any rotational angle desired by the viewer, anycombination of segments of the hepatic parenchy-ma, and portal and hepatic veins can be made toappear in varying levels of transparency. This featureis useful to visualize the extent of the hepatic veinsthrough the hepatic parenchyma and the arboriza-tion of the intrahepatic structures.

Our initial training course was designed to assessthe functionality and use of the model for surgicalresidents. A standard examination of 24 basicanatomic and relationship-function questions wasadministered to 6 senior general surgery residents.A short workshop (roughly 45 minutes) using theIHSEE was then given live with 1 instructor to 2physical locations joined only by the teleimmersivearchitecture. The standard examination was read-ministered immediately afterward and 6 monthslater. Thus, it was possible to evaluate the effective-ness of the IHSEE as an educational instrument.

RESULTSTraining on the IHSEE produced statistically sig-

nificant improvements in the test scores betweenthe pretests and posttests of the participants. Themean score increased from 17.67 to 23.67 (P < .02,

paired samples 2-tailed t test). We found no differ-ences between residents who were with the instruc-tor and those at the remote location. Five of 6residents wanted to take another workshop usingthis technology. Resident satisfaction interviewsconfirmed that the teleimmersive format was morelikely than conventional methods to help themvisualize the liver in the future. Delayed interviewswere particularly convincing; 1 resident reportedthat the workshop provided a new educational“framework” for understanding this complex mate-rial, allowing the resident to maximize understand-ing over the intervening months. This may explainwhy 6-month delayed testing demonstrated com-plete retention of new knowledge for all residentsand even substantial additional improvements forsome. See Fig 2 for presentation of raw data.

The learning objectives of the training coursewere divided into: (1) hepatic surface anatomy(questions in reference to the midplane of theliver: Cantlie’s line, Fissure of Ganz, umbilical fis-sure, Couinaud’s segment, and hilar plate), (2)Couinaud segmental anatomy (description anddemonstration of each segment), (3) surgicalresection nomenclature (IHPBA-standard resec-tion anatomy including first-order divisions such asright hemihepatectomy and second-order divisionssuch as various sectionectomies), and (4) intrahep-atic anatomy (description and demonstration ofright, middle, and left hepatic veins; descriptionand demonstration of right posterior/anterior andleft intrahepatic pedicles). No specific patternemerged in the residents’ improvement on thequestions. We believe this is consistent with ouroverall impression that they obtained a new broadunderstanding of the material.

Fig 2. Raw data from IHSEE pilot pretests, posttests, and delayed posttests.

Surgery Silverstein et al 277Volume 132, Number 2

DISCUSSIONThe high level of interest in this type of hepatic

visualization technology is apparent from the morethan 100 citations in the literature focused on 3Dliver anatomy. In the past decade the pace of noveltechnologies has outpaced the ability of medicineto incorporate these advancements. Imaging acqui-sition modalities have matured from early-genera-tion computed tomography scanners capable ofsimulated 3D reconstruction6 to the current daymultidetector spiral computed tomography andmagnetic resonance imaging. Parallel developmenthas occurred in fields using these modalities:anatomy,7 operative planning,8 and hepatic lesiontargeting.9

A limiting factor of simulated teaching aids andphysical models has been the lack of user interac-tivity, anatomic accuracy, and the limits of 2-dimen-sional materials, all of which are mitigated by therevolutionary visualization technology incorporat-ed into the IHSEE. Specifically, teleimmersive envi-ronments are an improvement over other teachingmethods because these environments provideinteractive stereo viewing of complex structuresfrom numerous real and theoretical vantage pointscorrectly and in a desirable way for the viewer with-out the need for biological materials. At the sametime, these environments permit expert educatorsto gain a wider audience than the few residents ontheir service for interactive discussion of advancedsurgical concepts and procedures. These featuresmay enable a new, more efficient, educationalframework that can extend across institutions.Thus, even if one could achieve the same improve-ment in test scores with conventional methods, theknowledge gained would require a tremendouseducator and resident investment in time andmaterials and would be nearly impossible to keepcurrent on an ongoing basis. Our preinterventiondata document the baseline test scores giventoday’s conventional educational methods: surgicalresidency. We show here that a single 45-minutesession by 1 master surgeon has the potential tocatapult and sustain understanding in multiple res-idents simultaneously.

By combining teleconferencing, telepresence,and virtual reality, these projection-based teleim-mersive systems have advantages over head-mount-ed and 2-dimensional displays because they allowparticipants freedom of motion plus the abilities tointeract with 3D models, point, gesture, converse,and see each other. Although the advent of multi-

media applications on compact disc and the cur-rent Internet have made it possible to extend thereach of instructional materials, we expect theadvent of teleimmersion to extend the reach ofinstruction itself. In fact, we suspect teleimmersivebiomedical environments for educational, clinical,and research collaborations will be so compellingthat they will be used even where collaborators arein the same location.

CONCLUSIONThis IHSEE provides an innovative interactive

educational framework enabling surgeons to over-come some of the barriers to teaching surgicalanatomic principles that require understanding ofcomplex 3D relationships. We have shown educa-tional efficacy of the IHSEE in a pilot study, whichsuggests that by using this environment, teachersand students could interact from locations world-wide to manipulate virtual anatomic informationand achieve their educational goals. Subsequentwork will include using the technology simultane-ously at multiple sites and piloting the use of simi-lar environments for clinical and researchcollaborations.

REFERENCES1. Strasberg SM, Belghiti J, Clavien PA, Gadzijev E, Garden JO,

Lau WY, et al. The Brisbane 2000 terminology of liver anato-my and resections. HPB Surg 2000;2:333-9.

2. Czernuszenko M, Pape D, Sandin D, DeFanti T, Dawe GL,Brown MD. The ImmersaDesk and Infinity Wall projection-based virtual reality displays. Comp Graph 1997;31:36-47.

3. Ackerman MJ. The Visible Human Project: a resource foreducation. Acad Med 1999;74:667-70.

4. Pearl RK, Evenhouse R, Rasmussen M, Dech F, Silverstein JC,Prokasy S, et al. The virtual pelvic floor, a teleimmersiveeducational environment. Proc AMIA Symp 1999;6:345-8.

5. Dech F, Ai Z, Silverstein JC. Manipulation of volumetricpatient data in a distributed virtual reality environment.Stud Health Technol Inform 2001;81:119-25.

6. Schlusselberg DS, Smith WK, Woodward DJ, Parkey RW.Use of computed tomography for a 3-dimensional treat-ment planning system. Comput Med Imaging Graph1988;12:25-32.

7. Soyer P, Bluemke DA, Bliss DF, Woodhouse CE, Fishman EK.Surgical segmental anatomy of the liver: demonstrationwith spiral CT during arterial portography and multiplanarreconstruction. AJR Am J Roentgenol 1994;163:99-103.

8. Glombitza G, Lamade W, Demiris AM, Gopfert MR, MayerA, Bahner ML, et al. Virtual planning of liver resections:image processing, visualization and volumetric evaluation.Int J Med Inf 1999;53:225-37.

9. Lamade W, Glombitza G, Fischer L, Chiu P, Cardenas CE,Thorn M. The impact of 3-dimensional reconstructions onoperation planning in liver surgery. Arch Surg 2000;135:1256-61.