cosmetic applications of virtual reality in aesthetic...

Cosmetic

Applications of Virtual Realityin Aesthetic SurgeryDarren M. Smith, M.D., Sherrell J. Aston, M.D., Court B. Cutting, M.D., and Aaron Oliker, M.S.Providence, R.I., and New York, N.Y.

Background: Virtual reality has a long his-tory in plastic and reconstructive surgery,with uses ranging from anatomical demon-stration to craniofacial surgical planning.The purpose of this article is to add to theliterature a computer graphics–based re-source for aesthetic surgery.Methods: Deformation tools, virtual cam-eras, and other components of Alias’s Maya4.0 were used to perform virtual surgicalprocedures on a detailed model of super-ficial facial anatomy. This three-dimen-sional model of superficial facial anatomy,derived from the National Library of Med-icine’s Visible Human Project, was also“aged” in Maya at key depths of anatomicaldissection. Adobe’s After Effects 5.5 wasused for animation postproduction workfor all animations.Results: Three-dimensional computer ani-mations were developed to illustrate tech-niques in aesthetic surgery. Another anima-tion was created that simulates facial aging atvarious levels of anatomical dissection.Conclusions: Computer modeling and ani-mation have the potential to play an impor-tant role in education, surgical planning, de-velopment, and other aspects of aestheticsurgery. (Plast. Reconstr. Surg. 116: 898,2005.)

Aesthetic surgery depends on an apprecia-tion of anatomy, surgical technique, and aes-

thetic judgment. Many tools exist for honingeach of these skills individually, but a mediumcapable of dynamically and simultaneously ex-ploring and teaching these components ofplastic surgery has yet to be developed. Thisstudy is a description of such a system designedto integrate anatomical accuracy and aestheticprinciples. We have described the benefits ofvirtual reality for teaching cleft lip–cleft palatesurgery in a previous study.1 This article focuseson the advantages of virtual reality germane toaesthetic surgery.

MATERIALS AND METHODS

Basic soft-tissue anatomical data were de-rived from the National Library of Medicine’sVisible Human Project2 (U.S. National Libraryof Medicine, Bethesda, Md.). These data werein the form of hematoxylin and eosin–stainedaxial histologic cuts of a female cadaver sec-tioned at 333-�m intervals. These cuts were thebasis for three-dimensional computer modelsgenerated using Alias’s Maya 4.0 (Alias SystemsCorp., Toronto, Ontario, Canada), accordingto a protocol we detailed in a previous study(Fig. 1).3 A skin model of the young femalehead was purchased commercially from theViewpoint Corporation (New York, N.Y.), andthe underlying soft-tissue models were manip-ulated within the limits of normal anatomy toconform to this skin shape.

Surgical procedures were animated in Mayausing a variety of techniques, some of which wedescribed in a previous study as they were ap-plied to simulating cleft palate surgery.1 Of

From Brown Medical School, the Institute of Reconstructive Plastic Surgery, New York University Medical Center, and the Plastic SurgeryDepartment of the Manhattan Eye, Ear, and Throat Hospital. Received for publication July 6, 2004; revised January 6, 2005.

Readers may also refer to the online version of the article at the Journal’sWeb site (www.plasreconsurg.org) for additional materials.

DOI: 10.1097/01.prs.0000176901.37684.8a

898

primary importance were Maya’s deformationtools, which we used to approximate the re-sponses of soft tissue to manipulations in eachstep of a surgical procedure. Virtual camerasand transparency also played a key role in theseanimations; together, they afforded uniqueperspectives on surgical technique.

We also used Maya to alter the topology ofthe skin model we purchased from Viewpoint.The resulting model was used to design a sys-tem to simulate the aging process of the femaleface, as described here. The skin model was“aged” in Maya using various deformation tech-niques to add periorbital ptosis, sagging of themalar fat pad, jowl formation, platysmal band-ing, and wrinkles following tension lines. Notethat because of technical considerations, wechose to age the skin model in Maya instead ofmorphing between laser scans of real faces atdifferent ages. The primary pitfall of the lattermethod is that laser scans of different faces atdifferent ages would necessarily have differenttopology (the data points defining each modelwould be different in order, location, andnumber). Inconsistent topology is not compat-ible with the morphing protocol used in Mayato age the skin model.

The models of deeper facial anatomy basedon the Visible Human Project described abovewere aged in Maya as well, to mirror thechanges occurring at the level of the skin.These modifications allowed us to create athree-dimensional animation that affords theviewer the capacity to dynamically visualize theaging process as it affects the three-dimen-sional skin model and underlying tissues. Fi-nally, Adobe’s After Effects 5.5 (Adobe Sys-

tems, Inc., San Jose, Calif.) was used foranimation postproduction work, includingcompositing layers of Maya footage and con-trolling the playback and sequencing of thisfootage.

RESULTS

Three animation sequences were completedin Maya. The first is an illustration of the fin-ger-assisted malar elevation developed by oneof us.4 The second demonstrates various tech-niques of superficial musculoaponeurotic sys-tem (SMAS) manipulation in rhytidectomy.These animation sequences serve as an exam-ple of how two specific surgical procedures canbe described with three-dimensional surgicalsimulation (Fig. 2).

The finger-assisted malar elevation anima-tion illustrates the subtleties of the manipula-tion of the malar fat pad in this procedure.Transparency and multiple camera angles wereused to illustrate the finger dissection in rela-tion to both the facial nerve and more super-ficial components of facial anatomy. The ani-mation illustrating various techniques of SMASalteration (movement and resection) in rhyti-dectomy illustrates both the SMAS manipula-tions themselves and the resultant effects atdifferent levels of anatomical dissection.

The third animation allows the user to ex-plore the aging process as it affects the face.Because the underlying anatomical modelswere manipulated to correspond to the exter-nal skin anatomy, one can view the aging faceat multiple levels of dissection. Also, the facial

FIG. 1. An anterolateral view of the model, with musclesand bony structures shown.

FIG. 2. A still image from the finger-assisted malar eleva-tion animation, just after the skin flap was raised. Here, theSMAS has been made invisible to portray the anatomy deepto the SMAS. The relationship of the facial nerve to thesurgeon’s anterior dissection is thus illustrated.

Vol. 116, No. 3 / APPLYING A VIRTUAL REALITY MODEL 899

model was designed such that any portion canbe made to age independently or in concertwith the rest of the face.

DISCUSSION

This project demonstrates the advantages ofvirtual reality methods in aesthetic surgery ed-ucation. We have sought to incorporate thestrengths of our previous virtual reality teach-ing endeavors into one suited to training infacial aesthetic surgery.1,5 Indeed, virtual realityhas been used as a teaching aid in manybranches of medicine: Colt et al. have exploredits uses in bronchoscopy training, Verma et al.have applied virtual reality to vitreoretinal sur-gery, Friedl et al. have used virtual reality forcardiac surgery, and Mabrey et al. and Pedow-itz et al. have discussed the role of virtual realtysimulation in arthroscopy, to name just a fewexamples from a wide field.6–10

The use of virtual reality has been describedbefore in plastic and reconstructive surgery.Craniofacial surgery was one of the first disci-plines in which computer modeling and virtualreality methods were used. The application ofcomputer graphics to craniofacial surgery has along history.11–29 Cutting et al. have used virtualreality methods intraoperatively to track bonefragment movement to a numerically opti-mized position.20 Our laboratory has designedvirtual reality animations to teach cleft palaterepair techniques.5 We have also developedanimations that illustrate the biomechanics ofeustachian tube dilation as it relates to cleftpalate repair.30

Our previous work has enumerated the ad-vantages of virtual reality in illustrating surgicaltechnique in general.1 Specific examples in-clude the use of “virtual cameras” to obtainanatomical views previously prevented by theminute size and inaccessible location of highlyrelevant anatomy, transparency to illustrate re-lationships between overlapping structures,and animation to describe dynamic processes.All of these techniques proved invaluable increating virtual reality representations of aes-thetic surgery’s techniques and most relevantanatomy. Specific applications of virtual realityto aesthetic surgery are described here.

Perspectives on Aging

The simulation of the aging skin provides aunique perspective on the process of aging.First, with simple manipulation of the Maya fileon which the aging simulations are based, one

is able to appreciate the details of the processfrom any angle and at any speed. This is usefulin understanding how the face as a whole agesand thus departs from the aesthetic ideals ofyouth. Moreover, one can watch one half of theface age while the other remains young for abasis of comparison. Individual components ofthe face can be aged, whereas others remainunchanged, allowing for an analysis of the im-pact each structure’s changing morphologywill have on the conformation of the face as awhole. Of course, these changes can be haltedat any intermediate point. In addition, any ofthese processes can be “rewound” together orin isolation to visualize the aging process inreverse. This capability represents an invalu-able way to define surgical goals: aesthetic sur-gery is not a direct reversal of the aging pro-cess; it is the best approximation of this endwith available technique. By offering a three-dimensional view of the aging process in re-verse, these animations can elucidate newpaths (and clarify existing ones) toward facialrejuvenation.

The aging process can be visualized not onlyfrom perspectives outside the skin but fromwithin the skin as well. The capacity to observethe model age from the inside out at variouslevels of dissection enhances the observer’s un-derstanding of the changes in underlying anat-omy that must occur beneath aging skin. Thesechanges are also simulated by virtual realityand provide a basis for an improved under-standing of the problems of aging to be solved.

Malar Fat Pad Manipulation in Finger-AssistedMalar Elevation

The malar fat pad is a central anatomicalplayer in the aging face. Our virtual realityanimation of the finger-assisted malar eleva-tion technique illustrates from a unique per-spective the subtleties entailed in manipulatingthis structure in relationship to surroundinganatomy. The finger-assisted malar elevationtechnique requires separation of the malar fatpad from underlying fascia and subsequent fatpad repositioning. Standard surgical video isincapable of providing a satisfactory view of themalar fat pad’s manipulation. It is impossible,with this medium, to view the malar fat pad inisolation or from an anterior perspective. Vir-tual reality makes such critical perspectivesavailable to the observer (Fig. 3; see also theonline animation).

900 PLASTIC AND RECONSTRUCTIVE SURGERY, September 1, 2005

Part of the difficulty of visualizing the malarfat pad arises from the fact that the fat pad isnot so much a discrete structure as a thicken-ing of the subcutaneous fat around the malarprominence. Virtual reality, however, allowedus to represent the malar fat pad as a discreettriangular structure to clarify its manipulationin the finger-assisted malar elevation proce-dure. The triangular representation of the fatpad is derived from the borders of the malarregion, where the subcutaneous fat is mostsignificantly thickened. These borders runroughly with the base of the triangle, resting onthe inferior border of the orbicularis oculi, thelateral leg following the course of the zygomat-icus major, and the medial leg, defining a siz-able portion of the nasolabial fold. Once themalar fat pad has been represented as a dis-crete structure, it is a trivial matter to view it inisolation (without surrounding, thinner, sub-cutaneous fat), allowing for clearer representa-tion of the finger-assisted malar elevation tech-nique.

Standard surgical video cannot capture ananterior intraoperative view of the malar fatpad simply because this pad is deep to andcontinuous with the skin of the malar regionand must remain thus positioned so that itsmovement in the finger-assisted malar eleva-tion procedure is transmitted to the overlyingskin. It is therefore impossible to see through

the malar skin from an anterior perspective,which is a substantial disadvantage to an ob-server attempting to understand the subtletiesof the procedure. Although two-dimensionalillustration allows for removal of the overlyingskin, this medium does not allow for the three-dimensional dynamic representation of surgi-cal maneuvers—another severe obstacle forone attempting to learn this three-dimension-ally subtle procedure. Three-dimensional ani-mation allows for direct and dynamic visualiza-tion of the malar fat pad manipulations fromany perspective, with or without surroundinganatomy.

Finally, the virtual reality models for the fin-ger-assisted malar elevation procedure are aprime example of the utility afforded by thecapacity to view the aging of the face from theinside out. This ability offers a useful methodof understanding the route the malar fat padfollows as aging occurs; the malar fat pad’saging process is better described as a gradualanterior-to-inferior flow than as a simple lineardescent. This distinction is very difficult to ar-ticulate with two-dimensional illustrations, andis important context for an understanding ofthe finger-assisted malar elevation technique.Although casually polled residents expressedthe opinion that the animations and modelswere helpful in enhancing their understandingof the procedures and anatomy described, aformal study measuring the technology’s use-fulness would be beneficial.

Force Vectors in SMAS Manipulation

The virtual reality model of facial anatomywas designed to enable one to demonstrate therange of SMAS force vectors used in varyingrhytidectomy techniques. These vectors arethree-dimensionally complex and cannot bedone justice by two-dimensional illustration.Surgical video is limited by practical consider-ations, including camera positioning, lighting,and obstructions that render these delicate vec-tor differences difficult to appreciate. Digitallighting and camera position promise a clearview of even the most subtle differences inoutcome following the application of variousforce vectors in SMAS manipulation. Thesevectors can be dynamically appreciated at anyanatomical depth. More than one depth can beshown simultaneously (Fig. 4). For example, byillustrating the application of the same forcevectors to both sides of the face, the skin effectson the right side of the face can be shown in

FIG. 3. This is another still image from the finger-assistedmalar elevation animation. The most cephalad portion of thehead is shown in its “wire frame” form, illustrating the typeof polygonal geometry on which all these models are based.The wire frame shown here is a modified version of the skinmodel acquired from Viewpoint. The skin is made invisible,allowing for direct visualization of the malar fat pad (withinthe green circle) and surrounding anatomy. The fat pad isshown partially transparent in this rendering to reveal theunderlying levators of the upper lip.


contrast with the deeper SMAS results on theleft. Underlying tissue manipulation is thus di-rectly compared with its superficial conse-quence from one half of the face to the other.

Planning Aesthetic Surgery

This medium not only permits new levels ofclarity in teaching surgical technique but mayalso address a long-standing conundrum in de-signing approaches to aesthetic surgery. Classi-cally, concepts of the aesthetic ideal existmerely as images of an individual’s most super-ficial anatomical surface: the skin. In reality,the visual impression of a face is the result ofthe complex interaction between structuresdeep to the skin and the skin itself. By conven-

tional methods of surgical planning, aestheticsurgery could only approximate these externalaesthetic standards indirectly by manipulatingunderlying structures. This is akin to hollowinga ball of clay, and attempting to create thelikeness of a face by modeling external featuresfrom the inner surface of the sphere.

Virtual reality aesthetic surgical simulationeliminates this cumbersome exercise of creat-ing a form in the image of its shadow. This endis achieved by allowing the user to selectivelyuncouple anatomical structures from their in-terdependence, promoting creative visualiza-tion. One might, for instance, change the su-perficial covering of the face (the skin) firstand plan the changes required to make the

FIG. 4. A four-panel illustration demonstrating the model’s capacity to show the effects of SMAS manipulationat different levels of dissection. Note the cervicomental angle, highlighted in green, in each panel. (Above, left) BeforeSMAS tension, superficial. (Above, right) Before SMAS tension, deep. (Below, left) After SMAS tension in the directionof the arrow, superficial. (Below, right) After SMAS tension in the direction of the arrow, deep. (Below) Note theincreased definition of the inferior mandibular border and the accentuation of the cervicomental angle. Animationin the online journal presents a clip from the finger-assisted malar elevation animation illustrating the fingerdissection. This segment demonstrates the elevation of the lateral border of the orbicularis oculi with the malar fatflap. The index finger is shown as it begins the blunt dissection deep to the orbicularis oculi and malar fat. Themalar fat is then made transparent to allow for clear visualization of the finger’s path inferior and medial as thedissection is continued to free the malar fat pad from its connection to the underlying fascia.


underlying structures fit the desired externalform second in a much more natural procedure.Moreover, once these internal adjustmentshave been idealized and the limitations posedby anatomical relationships encountered, thesurgeon can use virtual simulation to visualizethe most efficacious route of compromise be-tween structural requirements and the aes-thetic ideal.

Future Applications

The animations described in this study areonly the first of many we are currently devel-oping based on these models and techniques.Moreover, these models may in the future beincorporated into surgical simulators such asthat being developed by Cutting et al. allowingfor real-time haptic interaction between sur-geon and virtual reality structure.31 Such surgi-cal simulators will foster the design of novelaesthetic procedures. Another future goal isthe digitalization of a specific patient’s super-ficial anatomy with the aim of performing cus-tomized surgical planning based on these tech-niques and that individual’s anatomy, thusmoving far beyond today’s two-dimensionalsurgical outcome-predicting software.

CONCLUSIONS

This article presents a trio of three-dimen-sional computer animations on topics germaneto aesthetic surgery: the finger-assisted malarelevation method of rhytidectomy, compara-tive SMAS manipulations in rhytidectomy, andthe aging face. Interactive manipulation of theaging face simulation is also described. Theseresults demonstrate the potential uses of com-puter modeling and animation for aestheticsurgery, from education to planning.

Sherrell J. Aston, M.D.728 Park AvenueNew York, N.Y. [email protected]

REFERENCES

1. Cutting, C., Oliker, A., Haring, J., Dayan, J., and Smith,D. Use of three-dimensional computer graphic an-imation to illustrate cleft lip and palate surgery. Com-put. Aided Surg. 7: 326, 2002.

2 Peitgen, H. The Complete Visible Human: The Complete High-Resolution Male and Female Datasets from the Visible Hu-man Project. New York: Springer-Verlag, 1998.

3. Smith, D. M., Aston, S. J., Cutting, C. B., Oliker, A., andWeinzweig, J. Designing a virtual reality model foraesthetic surgery. Plast. Reconstr. Surg. 116: 83, 2005.

4. Aston, S. J. The ‘‘F.A.M.E.” facelift: Finger-assisted ma-lar elevation.In Proceedings of the The Cutting Edge:Aesthetic Surgery Symposium, New York, N.Y., 1998.

5. Cutting, C., Oliker, A., Harring, J., Dayan, J., and Smith,D. The Smile Train’s Virtual Surgery Videos: Cleft Surgery.New York: The Smile Train, 2002.

6. Colt, H. G., Crawford, S. W., and Galbraith, O., III. Vir-tual reality bronchoscopy simulation: A revolution inprocedural training. Chest 120: 1333, 2001.

7. Verma, D., Wills, D., and Verma, M. Virtual reality sim-ulator for vitreoretinal surgery. Eye 17: 71, 2003.

8. Friedl, R., Preisack, M. B., Klas, W., et al. Virtual realityand 3D visualizations in heart surgery education. HeartSurg. Forum 5: E17, 2002.

9. Mabrey, J. D., Gillogly, S. D., Kasser, J. R., et al. Virtualreality simulation of arthroscopy of the knee. Arthro-scopy 18: E28, 2002.

10. Pedowitz, R. A., Esch, J., and Snyder, S. Evaluation of avirtual reality simulator for arthroscopy skills devel-opment. Arthroscopy 18: E29, 2002.

11. Cutting, C., Bookstein, F., Grayson, B., and McCarthy,J. G. Computer aided planning of orthognathic sur-gery. In Proceedings of the Plastic Surgery ResearchCouncil, Detroit, Mich., 1984.

12. Cutting, C., Bookstein, F., Grayson, B., Fellingham, L.,and McCarthy, J. G. Three-dimensional computer-aided design of craniofacial surgical procedures. InD. Marchac (Ed.), Craniofacial Surgery . Berlin: Springer-Verlag, 1987. P. 17.

13. Cutting, C., Grayson, B., and Kim, H. Precision mul-tisegment bone positioning using computer aidedmethods in craniofacial surgical applications. In Pro-ceedings of the I.E.E.E. Engineering in Medicine andBiology Society, Philadelphia, Pa., November 1-4,1990.

14. Cutting, C. Applications of computer graphics to theevaluation and treatment of major craniofacial mal-formations. In J. U. a. G. Herman (Ed.), 3D Imaging inMedicine , Boca Raton, Fla.: CRC Press, 1991. Pp. 163–189.

15. Cutting, C., Taylor, R., and Bookstein, F. Comprehen-sive three dimensional cephalometric system for theplanning and execution of craniofacial surgical pro-cedures. In A. Montoya (Ed.), Craniofacial Surgery 4. InProceedings of the IV Biannual Meeting of the Inter-national Society of Craniomaxillofacial Surgery, inSantiago de Compostela, Spain, Monduzzi Editore,1991. Pp. 11–13.

16. Cutting, C., Taylor, R., and Bookstein, F. Computer-aided planning and execution of craniofacial surgicalprocedures. In Proceedings of the I.E.E.E. Engineer-ing in Medicine and Biology Society, Paris, France,October 29–November 1, 1992.

17. Cutting, C., Bookstein, F. L., Grayson, B., Fellingham, L.,and McCarthy, J. G. Three-dimensional computer-assisted design of craniofacial surgical procedures:Optimization and interaction with cephalometric andCT-based models. Plast. Reconstr. Surg. 77: 877, 1986.

18. Cutting, C., Dean, D., Bookstein, F. L., et al. A three-dimensional smooth surface analysis of untreatedCrouzon’s syndrome in the adult. J. Craniofac. Surg. 6:444, 1995.

19. Cutting, C., Grayson, B., Bookstein, F., Fellingham, L.,and McCarthy, J. G. Computer-aided planning andevaluation of facial and orthognathic surgery. Clin.Plast. Surg. 13: 449, 1986.


20. Cutting, C., Grayson, B., McCarthy, J. G., et al. A virtualreality system for bone fragment positioning in mul-tisegment craniofacial surgical procedures. Plast. Re-constr. Surg. 102: 2436, 1998.

21. Glat, P. M., Freund, R. M., Spector, J. A., et al. A clas-sification of plagiocephaly utilizing a three-dimen-sional computer analysis of cranial base landmarks.Ann. Plast. Surg. 36: 469, 1996.

22. Grayson, B., Cutting, C., Bookstein, F. L., Kim, H., andMcCarthy, J. G. The three-dimensional cephalo-gram: Theory, technique, and clinical application.Am. J. Orthod. Dentofacial Orthop. 94: 327, 1988.

23. Grayson, B. H., Cutting, C. B., Dufresne, C. R., Bookstein,F. L., McCarthy, J. G., and Patnaik, S. Three-dimen-sional computer simulation of craniofacial anatomy.N.Y. State Dent. J. 52: 29, 1986.

24. Lo, L. J., Marsh, J. L., Vannier, M. W., and Patel,V. V. Craniofacial computer-assisted surgical plan-ning and simulation. Clin. Plast. Surg. 21: 501, 1994.

25. Marsh, J. L., and Vannier, M. W. Surface imagingfrom computerized tomographic scans. Surgery 94:159, 1983.

26. Marsh, J. L., Vannier, M. W., Bresina, S., and Hemmer,K. M. Applications of computer graphics in cranio-facial surgery. Clin. Plast. Surg. 13: 441, 1986.

27. Marsh, J. L., Vannier, M. W., and Stevens, W. G. Surfacereconstructions from computerized tomographicscans for evaluation of malignant skull destruction.Am. J. Surg. 148: 530, 1984.

28. Marsh, J. L., Vannier, M. W., Stevens, W. G., Warren, J. O.,Gayou, D., and Dye, D. M. Computerized imagingfor soft tissue and osseous reconstruction in the headand neck. Clin. Plast. Surg. 12: 279, 1985.

29. Offutt, C. J., Vannier, M. W., Gilula, L. A., Marsh, J. L.,and Sutherland, C. J. Volumetric 3-D imaging ofcomputerized tomography scans. Radiol. Technol. 61:212, 1990.

30. Dayan, J., Smith, D., Cutting, C., Oliker, A., and Haring,J. A virtual reality model of eustachian tube dilationand clinical implications for cleft palate repair. Plast.Reconstr. Surg. 115: 236, 2005.

31. Cutting, C., Oliker, A., Khorammabadi, D., and Haddad,B. A deformer-based surgical simulator program forcleft lip and palate surgery. Submitted.


cosmetic applications of virtual reality in aesthetic...

Documents