- Email: [email protected]
Computer-simulated Eye Surgery
Computer.. simulated Eye Surgery A Novel Teaching Method for Residents and Practitioners Michael]. Sinclair, MS,l John W. Peifer, MA,l Ray Haleblian, 1 Malcolm N. Luxenberg, MD,Z Keith Green, PhD, DSc, 2•3 DavidS. Hull, MD2 Purpose: To describe an eye surgery simulator that uses a computerized graphic display to allow ophthalmic surgeons of all experience levels to enhance their surgical skills. Methods: The eye surgery simulation environment consists of a high-speed com puter graphics workstation, a stereo operating system, a wrist rest, and a position tracking stylus connected to force feedback motors. The surgeon views computer generated images of the eye and surgical instruments through the stereo operating system and controls the position and orientation of the chosen surgical instrument by moving the stylus. During the simulated instrument-tissue interactions, three feedback motors generate component force feedback along three orthogonal axes connected by thin rigid bars to the tip of the stylus. Results: The current proof-of-concept system provides a method for rapid learning experiences in a living eye simulation. Procedures can be recorded for playback and analysis, as well as for examination of techniques from different viewpoints (e.g., from inside the eye). Four simulated surgical instruments are available for use (scalpel, forceps, scissors, and phacoemulsifier). Conclusion: Eye surgery simulation offers both beginning and experienced ophthalmic surgeons an opportunity to learn new techniques and skills and achieve a satisfactory level of proficiency before use of that procedure in the operating room. When fully developed, this system should shorten the learning curve for new surgeons (i.e., residents) and offer an opportunity for practice before doing a difficult case or development of new techniques by experienced surgeons. The goal of replacement of current standard training methods for surgeons awaits further refinement and adjustment of the model. Ophthalmology 1995;102:517-521
Originally received: June 8, 1994. Revision accepted: September I, 1994. 1 Bioengineering Research Center, Georgia Institute of Technology, At lanta. 2 Department of Ophthalmology, Medical College of Georgia, Augusta. 3 Department of Physiology & Endocrinology, Medical College ofGeor gia, Augusta. Presented at the American Academy of Ophthalmology Annual Meeting, Chicago. October 1993, and at the Interactive Technology and Health Care meeting, San Diego, January 1994. Supported in part by an unrestricted departmental award to the De partment of Ophthalmology from Research to Prevent Blindness, Inc, New York, New York, and in part by an MCG/GIT Biomedical Research and Education Program. Dr. Green is the recipient of a Senior Scientific Investigator Award from Research to Prevent Blindness, Inc, New York, New York.
Computer-driven flight simulators have been designed to provide an interactive, real-time experience that realisti cally simulates emergency events for the experienced pilot, offers exposure to new environments (i.e., practice land ings at airports new to the pilot), provides fulfillment of recertification requirements, or offers learning experiences for the trainee. 1 The technology required to produce re alistic visual simulation previously has required extremely expensive custom-built simulation computers, but now off-the-shelf, high-performance graphics workstations The authors do not have any commercial or proprietary interest in any of the companies or products mentioned in this article. Reprint requests to Keith Green, PhD, DSc, Department of Ophthal mology, Medical College of Georgia, Augusta, GA 30912-3400.
517
Ophthalmology
Volume 102, Number 3, March 1995
make the generation and presentation of realistic simu lation imagery in real-time economically possible. Al though the current cost of computer hardware to run the appropriate software is expensive, based on the history of computer development it is likely that the cost of the hardware will be reduced significantly in the future. Com puter simulations, especially those where real-time visuals are presented as an outcome of a person-in-the-loop in teraction, have established their significance in training. 1 Ophthalmic surgery is an area that can benefit from such technology. Historically, physicians learn new sur gical procedures through observation, practicing on ani mals and/or donor eyes, and then assisting with, or per forming, the procedure on patients under the supervision of an experienced surgeon. Studies have shown that oph thalmology residents can perform cataract-intraocular lens surgery, as well as other procedures, at a safety and efficacy level comparable to community professional standards when given appropriate training, education, and supervision. 2 - 14 In the future, it is anticipated that com puter simulation will allow physicians to practice surgical procedures and hone skills in a virtual environment in which (1) there is no risk to patients, 15 (2) any errors can be reset immediately by the computer, and (3) the pro cedure can be reviewed from new, insightful perspectives that are impossible in the real world (e.g., from inside the eye).
Materials and Methods Eye surgery simulation includes five basic components that are required to support surgical training: ( 1) an op erating station, (2) computer models of the anatomy and the surgical instruments, (3) a tactile feedback system, (4) a position tracking system for the surgical instruments, and (5) software that controls the interaction and updates the visual and tactile feedback. The operating station is the physical interface between the surgeon and the virtual operating environment. The operating station includes a stereo viewing system and a wrist rest, but instead oflooking directly at a real eye, the surgeon interacts with a virtual eye using a virtual surgical instrument controlled by a hand-held Polhemus three dimensional position tracking stylus (Polhemus, Inc, Colchester, VT) that continuously reports position and orientation to the computer (Fig 1). The tip of the stylus is connected by thin rigid bars to three motors that gen erate component force feedback in response to the surgical interaction. The simulation interface also includes a dial box and a button box that control instrument selection and viewing options. Using the button box, training ses sions can be changed, recorded, and played back. In ad dition, the models may be reset and outer layers of the eye can be removed to show interior anatomic compo nents. Dials allow rotation of the model so that different views may be observed: changes may be made in the transparency of tissues such as the cornea or lens, the degree of magnification may be changed using a zoom system, and adjustment of stereo viewing parameters may
518
be modified. An instrument activation switch on the stylus controls instrument actions such as opening and closing forceps and scissors. These actions occur instantly upon activation. The software rapidly cycles through a loop that reads the current instrument status and activates the appropriate graphic and tactile responses. The simulation latency, de fined as the period of time between an operator input (stylus movement) and the last pixel of the first field drawn (which reflects that movement of the stylus) is approxi mately 120 mseconds. Tactile feedback, however, is ap proximately 300 11seconds because no visual feedback is involved in this aspect. Instrument status includes the type of surgical instrument, whether it is activated (for example, scissors open or closed), and its position and orientation. Four surgical instruments have been modeled: (1) a scalpel, which cuts at the point of application; (2) scissors, which open and close and cut tissue at the point of application; (3) forceps that open and close, allowing one to grasp and pull at the point of application; and (4) a phacoemulsifier that removes the lens (Fig 2). In the virtual environment, the eye and the surgical instruments exist only as computer models. The eye is represented by a collection of deformable three-dimen sional polygonal surface models. Surface models of the sclera, iris, zonules, and retina are texture-mapped with photographic images of these components. The lens and cornea are modeled as semi-transparent surfaces. The graphic images required in the simulation were obtained from videotapes and photographs. The geometric, me chanical, and textural components of the ocular structures of interest were obtained from ocular models and real specimens. Measurements were made of the forces in volved in cutting the sclera using strain gauges attached to scalpels that were used to cut into eye bank eyes. Ad justments were made in the simulation to reflect tissue behavior and deformability, as described by experienced surgeons, as accurately as possible at this level of devel opment. The four surgical instruments used in the simulation were created as simple three-dimensional polygonal sur face models using the Wavefront Advanced Visualizer modeling package (Wavefront Technologies, Santa Bar bara, CA). An overhead light source included in the sim ulation produces specular highlights on the ocular com ponents. The texture mapping of photographic images, transparency, and lighting are supported by the Silicon Graphics system (Silicon Graphics, Mountain View, CA). These advanced computer graphics techniques add a high degree of realism to the anatomic models that could not have been achieved in previous generations ofcomputers.
Results Interaction between the instruments and the eye depends on the instrument per se, the location of the instrument in the anatomy, and the kind of action requested by the surgeon. As the knife makes contact with the external ocular surface, the sclera slightly deforms until the blade
Sinclair et al · Computer Graphic Eye Surgery Simulation
Force Feedback
0
Figure 1. Diagram shows dif ferent components of the simulation system. The op erator sits at a wrist support system and views the virtual eye via a stereo eyepiece. Movement of the hand-held stylus causes force and posi tion feedback that is displayed in the movement of the vir tual instrument on the virtual eye. All movements and de sired ocular images are trans lated into action by the graphics workstation.
Simulation Computer I Graphics Workstation penetrates and starts to cut. The force (or tactile) feedback system produces a compliant resistance as the sclera de forms and then allows the blade to slice through the sclera with a small viscous resistance in the cutting direction after penetration (Fig 2). As the blade is cutting, a strong compliant resistance is generated at the stylus tip in the direction orthogonal (lateral) to the cutting direction to produce the same type of resistance that would be expe rienced if the surgeon tried to lift the incision with the flat portion of the blade. Surgical scissors function similarly to the knife; how ever, the scissors close and open when the stylus activation switch is depressed and released, respectively (Fig 2). After an incision is made, the scleral surface is redefined to include a wound where the cut was made. Forceps can be used to grasp the sclera and open the wound (Fig 2). During the grasping action, the simulation software con tinuously recomputes the deformed surface, based on how far the sclera is stretched. Both visual and force feedback are provided to the user: if the forceps are opened during the tissue-grasping action, the sclera is released, it reverts to its original shape visually, and the resistance is removed. The phacoemulsifier can be inserted through the wound and directed to the lens. During this procedure, resistance is produced to hold the phacoemulsifier in the wound, and another response is produced to simulate contact with the lens. By pressing the instrument activation switch when the phacoemulsifier is in contact with the lens, the resistance changes to viscous to simulate the feel of the vacuum action as the phacoemulsifier removes the clouded lens (Fig 2). A computer model thus has been developed that is functional and offers features desirable and necessary in such a simulation of ocular surgical procedures. Further work is necessary to enhance different features of the model system, including refinement of the tactile feed
back, addition of a second instrument in the operating field, adjustments of the anatomy, elimination of slight time lag between stylus movement and instrument move ment in the visual feedback system, and more realistic simulation of the force involved in lens removal. During this next development phase, both novice and trained surgeons will provide feedback on the extent to which the simulation duplicates the surgical experience. Currently, trained surgeons have had only limited use of the model system.
Discussion Use of the eye surgery simulator would permit the ex pertise of skilled surgeons to be used to educate less-ex perienced surgeons, allowing them to learn and practice basic skills for any surgical procedure while being evalu ated objectively by an instructor. Similarly, the oppor tunity exists to simulate infrequently performed surgical situations for experienced surgeons, so that management of these can be practiced. The simulator allows review upon its conclusion, and an idealized method of perform ing the procedure also can be demonstrated. The entire eye can be modeled with the appropriate structures and viewed from any desired angle. This approach will de crease the need for either animal or human donor eyes as the methodology becomes more fully developed. Simulation offers great potential for improving medical care through training, instrument design, surgical plan ning, and support. Residents will use simulators to explore basic anatomy and practice hand-eye coordination for many procedures that currently require training with an imal and/or donor eyes. Advanced training will be pro vided on simulators for experienced surgeons as they try new procedures or experiment with different instru
519
Ophthalmology
Volume 102, Number 3, March 1995
Figure 2. Photographs of computer-generated images. A, a scleral incision; B, another view of a scleral incision; C, scissors are used to expand the wound (the wound is large for illustrative purposes only, and is not designed to replicate any specific surgical process); D, forceps are used to pull the sclera away from the wound (no conjunctiva currently is included in the model, and the deformation shown is to illustrate the nature of the type of deformation, not one that necessarily occurs during any specific surgical procedure); E, insertion of phacoemulsifier into the eye; F, phacoemulsifier removes lens.
ments. 15 When fully developed, therefore, it will allow residents to develop surgical techniques in a harm-proof environment and will allow practitioners to acquire new techniques without compromising patient care. Anatomy can be peeled away, one layer at a time, to show relations among the instruments, the incision, and critical ocular structures. With patient-specific models of the anatomy, surgery could be simulated before it actually is performed on the patient. Optimum instrumentation, approach an gle, and depth of incision can be predetermined through practice on the simulator. The use of computer technology in education is being used increasingly in different fields of medicine, extending from laparoscopic surgery to anesthesia. 16 In the latter, it is being used to improve the management of critical sit uations.17·18 Reconstructive craniofacial surgery also has benefited from the use of surgical simulation based on three-dimensional imaging to display different possible surface configurations after simulated surgical manipu lation of the underlying bones. 19·20 An increase in mor
520
bidity and mortality has been observed when unskilled practitioners first begin performing new procedures, 16·21 thereby stressing the importance of providing some means by which inexperienced, and experienced, surgeons can learn new techniques. Many new procedures are complex and require the acquisition of considerable skills. By pro viding simulation as a facet of the learning process, an important step in training is achieved. Use ofadvanced technology with computer simulators of laparoscopic techniques also is being proposed as a facet of privileging criteria, although this requires the sur geon to move past the initial learning curve and become somewhat proficient with the techniques. 22 The problems to be addressed concerning this use of simulation tech nology are those inherent in the adoption of any new techniques, namely the establishment of guidelines for credentialing or certifying physicians to perform these procedures. Although use of simulation is common in the airline industry for training pilots and measuring skill lev els, 1 its introduction into medicine will require careful consideration. These approaches will not fully replace laboratory-acquired skills, procedure monitoring, and as sociation with those more skilled in a given procedure. The development and adaptation of simulator technology would allow assessment of the technical demands of a new procedure and measurement of a physician's skills against standardized models. 22 Thus, simulation offers not only a training methodology but also may offer an eval uative technology where accurate, objective assessment of technical skills may be made.
References I. Haber RN. Right simulation. Sci Am 1986;255:96-103. 2. Jaffe NS. The changing scene of intraocular implant lens surgery. Am J Ophthalmol 1979;88:819-28. 3. Kersten RC, Kolder HE. Intraocular lens implantation: res idents vs. staff. Ophthalmic Surg 1982;13:470-4. 4. Sutton GA. Intraocular lens implantation by surgeons in training. Br J Ophthalmol 1980;64:687-8. 5. Neuhaus RW, Straatsma BR. Cataract surgery with intra ocular lens implantation by ophthalmology resident sur geons. Ophthalmic Surg 1983;14:415-7. 6. Schanzer MC, Wilhelmus KR. Outpatient cataract surgery by ophthalmology residents in a county hospital. Ann Ophthalmol 1985; 17:480-2. 7. Kirby TJ. A course in surgical technique for residents in oph thalmology. Trans Am Ophthalmol Soc 1983;81:333-57. 8. Prince RB, Tax RL, Miller DH. Conversion to small-incision phacoemulsification: experience with the first 50 eyes. J Cataract Refract Surg 1993;19:246-50. 9. Trakkanen A, Eskelin L. Training in ophthalmic surgery. The results of the first five cataract extractions of 60 ophthalmic residents from 1944 to 1966. Int Surg 1968;49: 342-7. 10. Browning DJ, Cobo LM. Early experience in extracapsular cat aract surgery by residents. Ophthalmology 1985;92:1647-53. II. Pearson PA, Owen DG, Van Meter WS, Smith TJ. Vitreous loss rates in extracapsular cataract surgery by residents. Ophthalmology 1989;96: 1225-7. 12. Straatsma BR, Meyer KT, Bastek JV, Lightfoot DO. Pos terior chamber intraocular lens implantation by ophthal
Sinclair et al · Computer Graphic Eye Surgery Simulation
13.
14. 15. 16. 17.
mology residents. A prospective study of cataract surgery. Ophthalmology 1983;90:327-35. Cruz OA, Wallace GW, Gay CA, et al. Visual results and complications ofphacoemulsification with intraocular lens implantation performed by ophthalmology residents. Oph thalmology 1992;99:448-52. Allinson RW, Metrikin DC, Fante RG. Incidence of vitreous loss among third-year residents performing phacoemulsifi cation. Ophthalmology 1992;99:726-30. Satava RM. Virtual reality surgical simulator. The first steps. Surg Endosc 1993;7:203-5. Grundfest WS. Credentialing in an era ofchange [editorial]. JAMA 1993;270:2725. Schwid HA, O'Donnell D. Anesthesiologists' management of simulated critical incidents. Anesthesiology 1992;76:495 501.
18. Gaba OM. Improving anesthesiologists' performance by simulating reality [editorial]. Anesthesiology 1992;76:491 4. 19. Zonneveld FW, Lobregt S, van der Meulen JC, Vaandrager JM. Three-dimensional imaging in craniofacial surgery. World J Surg 1989;13:328-42. 20. Fujioka M, Yokoi S, Yasuda Y, et al. Computer-aided in teractive surgical simulation for craniofacial anomalies based on 3-D surface reconstruction CT images. Radiat Med 1988;6:204-12. 21. Sharp WV, Guyton DP, Crans CA, et al. Initial experience with laparoscopic surgery: establishing a new surgical pro cedure. J Laparoendosc Surg 1992;2: 151-5. 22. Dent TL. Training, credentialing, and granting of clinical privileges for laparoscopic general surgery. Am J Surg 1991; 161:399-403.
521
Originally received: June 8, 1994. Revision accepted: September I, 1994. 1 Bioengineering Research Center, Georgia Institute of Technology, At lanta. 2 Department of Ophthalmology, Medical College of Georgia, Augusta. 3 Department of Physiology & Endocrinology, Medical College ofGeor gia, Augusta. Presented at the American Academy of Ophthalmology Annual Meeting, Chicago. October 1993, and at the Interactive Technology and Health Care meeting, San Diego, January 1994. Supported in part by an unrestricted departmental award to the De partment of Ophthalmology from Research to Prevent Blindness, Inc, New York, New York, and in part by an MCG/GIT Biomedical Research and Education Program. Dr. Green is the recipient of a Senior Scientific Investigator Award from Research to Prevent Blindness, Inc, New York, New York.
Computer-driven flight simulators have been designed to provide an interactive, real-time experience that realisti cally simulates emergency events for the experienced pilot, offers exposure to new environments (i.e., practice land ings at airports new to the pilot), provides fulfillment of recertification requirements, or offers learning experiences for the trainee. 1 The technology required to produce re alistic visual simulation previously has required extremely expensive custom-built simulation computers, but now off-the-shelf, high-performance graphics workstations The authors do not have any commercial or proprietary interest in any of the companies or products mentioned in this article. Reprint requests to Keith Green, PhD, DSc, Department of Ophthal mology, Medical College of Georgia, Augusta, GA 30912-3400.
517
Ophthalmology
Volume 102, Number 3, March 1995
make the generation and presentation of realistic simu lation imagery in real-time economically possible. Al though the current cost of computer hardware to run the appropriate software is expensive, based on the history of computer development it is likely that the cost of the hardware will be reduced significantly in the future. Com puter simulations, especially those where real-time visuals are presented as an outcome of a person-in-the-loop in teraction, have established their significance in training. 1 Ophthalmic surgery is an area that can benefit from such technology. Historically, physicians learn new sur gical procedures through observation, practicing on ani mals and/or donor eyes, and then assisting with, or per forming, the procedure on patients under the supervision of an experienced surgeon. Studies have shown that oph thalmology residents can perform cataract-intraocular lens surgery, as well as other procedures, at a safety and efficacy level comparable to community professional standards when given appropriate training, education, and supervision. 2 - 14 In the future, it is anticipated that com puter simulation will allow physicians to practice surgical procedures and hone skills in a virtual environment in which (1) there is no risk to patients, 15 (2) any errors can be reset immediately by the computer, and (3) the pro cedure can be reviewed from new, insightful perspectives that are impossible in the real world (e.g., from inside the eye).
Materials and Methods Eye surgery simulation includes five basic components that are required to support surgical training: ( 1) an op erating station, (2) computer models of the anatomy and the surgical instruments, (3) a tactile feedback system, (4) a position tracking system for the surgical instruments, and (5) software that controls the interaction and updates the visual and tactile feedback. The operating station is the physical interface between the surgeon and the virtual operating environment. The operating station includes a stereo viewing system and a wrist rest, but instead oflooking directly at a real eye, the surgeon interacts with a virtual eye using a virtual surgical instrument controlled by a hand-held Polhemus three dimensional position tracking stylus (Polhemus, Inc, Colchester, VT) that continuously reports position and orientation to the computer (Fig 1). The tip of the stylus is connected by thin rigid bars to three motors that gen erate component force feedback in response to the surgical interaction. The simulation interface also includes a dial box and a button box that control instrument selection and viewing options. Using the button box, training ses sions can be changed, recorded, and played back. In ad dition, the models may be reset and outer layers of the eye can be removed to show interior anatomic compo nents. Dials allow rotation of the model so that different views may be observed: changes may be made in the transparency of tissues such as the cornea or lens, the degree of magnification may be changed using a zoom system, and adjustment of stereo viewing parameters may
518
be modified. An instrument activation switch on the stylus controls instrument actions such as opening and closing forceps and scissors. These actions occur instantly upon activation. The software rapidly cycles through a loop that reads the current instrument status and activates the appropriate graphic and tactile responses. The simulation latency, de fined as the period of time between an operator input (stylus movement) and the last pixel of the first field drawn (which reflects that movement of the stylus) is approxi mately 120 mseconds. Tactile feedback, however, is ap proximately 300 11seconds because no visual feedback is involved in this aspect. Instrument status includes the type of surgical instrument, whether it is activated (for example, scissors open or closed), and its position and orientation. Four surgical instruments have been modeled: (1) a scalpel, which cuts at the point of application; (2) scissors, which open and close and cut tissue at the point of application; (3) forceps that open and close, allowing one to grasp and pull at the point of application; and (4) a phacoemulsifier that removes the lens (Fig 2). In the virtual environment, the eye and the surgical instruments exist only as computer models. The eye is represented by a collection of deformable three-dimen sional polygonal surface models. Surface models of the sclera, iris, zonules, and retina are texture-mapped with photographic images of these components. The lens and cornea are modeled as semi-transparent surfaces. The graphic images required in the simulation were obtained from videotapes and photographs. The geometric, me chanical, and textural components of the ocular structures of interest were obtained from ocular models and real specimens. Measurements were made of the forces in volved in cutting the sclera using strain gauges attached to scalpels that were used to cut into eye bank eyes. Ad justments were made in the simulation to reflect tissue behavior and deformability, as described by experienced surgeons, as accurately as possible at this level of devel opment. The four surgical instruments used in the simulation were created as simple three-dimensional polygonal sur face models using the Wavefront Advanced Visualizer modeling package (Wavefront Technologies, Santa Bar bara, CA). An overhead light source included in the sim ulation produces specular highlights on the ocular com ponents. The texture mapping of photographic images, transparency, and lighting are supported by the Silicon Graphics system (Silicon Graphics, Mountain View, CA). These advanced computer graphics techniques add a high degree of realism to the anatomic models that could not have been achieved in previous generations ofcomputers.
Results Interaction between the instruments and the eye depends on the instrument per se, the location of the instrument in the anatomy, and the kind of action requested by the surgeon. As the knife makes contact with the external ocular surface, the sclera slightly deforms until the blade
Sinclair et al · Computer Graphic Eye Surgery Simulation
Force Feedback
0
Figure 1. Diagram shows dif ferent components of the simulation system. The op erator sits at a wrist support system and views the virtual eye via a stereo eyepiece. Movement of the hand-held stylus causes force and posi tion feedback that is displayed in the movement of the vir tual instrument on the virtual eye. All movements and de sired ocular images are trans lated into action by the graphics workstation.
Simulation Computer I Graphics Workstation penetrates and starts to cut. The force (or tactile) feedback system produces a compliant resistance as the sclera de forms and then allows the blade to slice through the sclera with a small viscous resistance in the cutting direction after penetration (Fig 2). As the blade is cutting, a strong compliant resistance is generated at the stylus tip in the direction orthogonal (lateral) to the cutting direction to produce the same type of resistance that would be expe rienced if the surgeon tried to lift the incision with the flat portion of the blade. Surgical scissors function similarly to the knife; how ever, the scissors close and open when the stylus activation switch is depressed and released, respectively (Fig 2). After an incision is made, the scleral surface is redefined to include a wound where the cut was made. Forceps can be used to grasp the sclera and open the wound (Fig 2). During the grasping action, the simulation software con tinuously recomputes the deformed surface, based on how far the sclera is stretched. Both visual and force feedback are provided to the user: if the forceps are opened during the tissue-grasping action, the sclera is released, it reverts to its original shape visually, and the resistance is removed. The phacoemulsifier can be inserted through the wound and directed to the lens. During this procedure, resistance is produced to hold the phacoemulsifier in the wound, and another response is produced to simulate contact with the lens. By pressing the instrument activation switch when the phacoemulsifier is in contact with the lens, the resistance changes to viscous to simulate the feel of the vacuum action as the phacoemulsifier removes the clouded lens (Fig 2). A computer model thus has been developed that is functional and offers features desirable and necessary in such a simulation of ocular surgical procedures. Further work is necessary to enhance different features of the model system, including refinement of the tactile feed
back, addition of a second instrument in the operating field, adjustments of the anatomy, elimination of slight time lag between stylus movement and instrument move ment in the visual feedback system, and more realistic simulation of the force involved in lens removal. During this next development phase, both novice and trained surgeons will provide feedback on the extent to which the simulation duplicates the surgical experience. Currently, trained surgeons have had only limited use of the model system.
Discussion Use of the eye surgery simulator would permit the ex pertise of skilled surgeons to be used to educate less-ex perienced surgeons, allowing them to learn and practice basic skills for any surgical procedure while being evalu ated objectively by an instructor. Similarly, the oppor tunity exists to simulate infrequently performed surgical situations for experienced surgeons, so that management of these can be practiced. The simulator allows review upon its conclusion, and an idealized method of perform ing the procedure also can be demonstrated. The entire eye can be modeled with the appropriate structures and viewed from any desired angle. This approach will de crease the need for either animal or human donor eyes as the methodology becomes more fully developed. Simulation offers great potential for improving medical care through training, instrument design, surgical plan ning, and support. Residents will use simulators to explore basic anatomy and practice hand-eye coordination for many procedures that currently require training with an imal and/or donor eyes. Advanced training will be pro vided on simulators for experienced surgeons as they try new procedures or experiment with different instru
519
Ophthalmology
Volume 102, Number 3, March 1995
Figure 2. Photographs of computer-generated images. A, a scleral incision; B, another view of a scleral incision; C, scissors are used to expand the wound (the wound is large for illustrative purposes only, and is not designed to replicate any specific surgical process); D, forceps are used to pull the sclera away from the wound (no conjunctiva currently is included in the model, and the deformation shown is to illustrate the nature of the type of deformation, not one that necessarily occurs during any specific surgical procedure); E, insertion of phacoemulsifier into the eye; F, phacoemulsifier removes lens.
ments. 15 When fully developed, therefore, it will allow residents to develop surgical techniques in a harm-proof environment and will allow practitioners to acquire new techniques without compromising patient care. Anatomy can be peeled away, one layer at a time, to show relations among the instruments, the incision, and critical ocular structures. With patient-specific models of the anatomy, surgery could be simulated before it actually is performed on the patient. Optimum instrumentation, approach an gle, and depth of incision can be predetermined through practice on the simulator. The use of computer technology in education is being used increasingly in different fields of medicine, extending from laparoscopic surgery to anesthesia. 16 In the latter, it is being used to improve the management of critical sit uations.17·18 Reconstructive craniofacial surgery also has benefited from the use of surgical simulation based on three-dimensional imaging to display different possible surface configurations after simulated surgical manipu lation of the underlying bones. 19·20 An increase in mor
520
bidity and mortality has been observed when unskilled practitioners first begin performing new procedures, 16·21 thereby stressing the importance of providing some means by which inexperienced, and experienced, surgeons can learn new techniques. Many new procedures are complex and require the acquisition of considerable skills. By pro viding simulation as a facet of the learning process, an important step in training is achieved. Use ofadvanced technology with computer simulators of laparoscopic techniques also is being proposed as a facet of privileging criteria, although this requires the sur geon to move past the initial learning curve and become somewhat proficient with the techniques. 22 The problems to be addressed concerning this use of simulation tech nology are those inherent in the adoption of any new techniques, namely the establishment of guidelines for credentialing or certifying physicians to perform these procedures. Although use of simulation is common in the airline industry for training pilots and measuring skill lev els, 1 its introduction into medicine will require careful consideration. These approaches will not fully replace laboratory-acquired skills, procedure monitoring, and as sociation with those more skilled in a given procedure. The development and adaptation of simulator technology would allow assessment of the technical demands of a new procedure and measurement of a physician's skills against standardized models. 22 Thus, simulation offers not only a training methodology but also may offer an eval uative technology where accurate, objective assessment of technical skills may be made.
References I. Haber RN. Right simulation. Sci Am 1986;255:96-103. 2. Jaffe NS. The changing scene of intraocular implant lens surgery. Am J Ophthalmol 1979;88:819-28. 3. Kersten RC, Kolder HE. Intraocular lens implantation: res idents vs. staff. Ophthalmic Surg 1982;13:470-4. 4. Sutton GA. Intraocular lens implantation by surgeons in training. Br J Ophthalmol 1980;64:687-8. 5. Neuhaus RW, Straatsma BR. Cataract surgery with intra ocular lens implantation by ophthalmology resident sur geons. Ophthalmic Surg 1983;14:415-7. 6. Schanzer MC, Wilhelmus KR. Outpatient cataract surgery by ophthalmology residents in a county hospital. Ann Ophthalmol 1985; 17:480-2. 7. Kirby TJ. A course in surgical technique for residents in oph thalmology. Trans Am Ophthalmol Soc 1983;81:333-57. 8. Prince RB, Tax RL, Miller DH. Conversion to small-incision phacoemulsification: experience with the first 50 eyes. J Cataract Refract Surg 1993;19:246-50. 9. Trakkanen A, Eskelin L. Training in ophthalmic surgery. The results of the first five cataract extractions of 60 ophthalmic residents from 1944 to 1966. Int Surg 1968;49: 342-7. 10. Browning DJ, Cobo LM. Early experience in extracapsular cat aract surgery by residents. Ophthalmology 1985;92:1647-53. II. Pearson PA, Owen DG, Van Meter WS, Smith TJ. Vitreous loss rates in extracapsular cataract surgery by residents. Ophthalmology 1989;96: 1225-7. 12. Straatsma BR, Meyer KT, Bastek JV, Lightfoot DO. Pos terior chamber intraocular lens implantation by ophthal
Sinclair et al · Computer Graphic Eye Surgery Simulation
13.
14. 15. 16. 17.
mology residents. A prospective study of cataract surgery. Ophthalmology 1983;90:327-35. Cruz OA, Wallace GW, Gay CA, et al. Visual results and complications ofphacoemulsification with intraocular lens implantation performed by ophthalmology residents. Oph thalmology 1992;99:448-52. Allinson RW, Metrikin DC, Fante RG. Incidence of vitreous loss among third-year residents performing phacoemulsifi cation. Ophthalmology 1992;99:726-30. Satava RM. Virtual reality surgical simulator. The first steps. Surg Endosc 1993;7:203-5. Grundfest WS. Credentialing in an era ofchange [editorial]. JAMA 1993;270:2725. Schwid HA, O'Donnell D. Anesthesiologists' management of simulated critical incidents. Anesthesiology 1992;76:495 501.
18. Gaba OM. Improving anesthesiologists' performance by simulating reality [editorial]. Anesthesiology 1992;76:491 4. 19. Zonneveld FW, Lobregt S, van der Meulen JC, Vaandrager JM. Three-dimensional imaging in craniofacial surgery. World J Surg 1989;13:328-42. 20. Fujioka M, Yokoi S, Yasuda Y, et al. Computer-aided in teractive surgical simulation for craniofacial anomalies based on 3-D surface reconstruction CT images. Radiat Med 1988;6:204-12. 21. Sharp WV, Guyton DP, Crans CA, et al. Initial experience with laparoscopic surgery: establishing a new surgical pro cedure. J Laparoendosc Surg 1992;2: 151-5. 22. Dent TL. Training, credentialing, and granting of clinical privileges for laparoscopic general surgery. Am J Surg 1991; 161:399-403.
521