Surg Endosc (2010) 24:1341–1346 DOI 10.1007/s00464-009-0771-1
Adaptation to a dynamic visual perspective in laparoscopy through training in the cutting task Arturo Minor Martı´nez • Jose´ Luis Limo´n Aguilar Ricardo Ordorica Flores • Jose´ Luis Ortiz Simo´n • Alejandro Garcı´a Pe´rez
•
Received: 1 July 2009 / Accepted: 11 November 2009 / Published online: 24 December 2009 Ó Springer Science+Business Media, LLC 2009
Abstract Background Laparoscopic surgery demands of surgeons special skills acquired only through practice. Laparoscopic training systems traditionally have an optical system that, once positioned, remains fixed and cannot refresh the perspective unless the task is interrupted and the camera repositioned. During a surgery, the visual perspective changes constantly to relocate the surgical target. This difference is a limitation for any novice surgeon. This report proposes the use of a mechatronic system that allows the trainee to handle optics dynamically during training in the cutting task and thereby adapt to dynamic relocation of the surgical target. Methods The study was conducted in two phases. The first phase involved using fixed optics to cut a circle drawn on a piece of cloth. The second phase involved the same cutting task but with the visual perspective changed dynamically by the user via a mechatronic assistant. Results The data show that by adapting to dynamic optics, medical trainees can quickly and easily handle and locate the task with real-time changes in visual perspective and can also improve task quality. A significant statistical
A. M. Martı´nez (&) J. L. O. Simo´n A. G. Pe´rez Bioelectronics Section, Centro de Investigacio´n y Estudios Avanzados IPN, Av. IPN 2508, C.P. 073000 Mexico City, Mexico e-mail:
[email protected] J. L. O. Simo´n e-mail:
[email protected] J. L. L. Aguilar ´ ngeles, Clı´nica Londres, Mexico City, Mexico Hospital A R. O. Flores Hospital Infantil, Federico Go´mez, Mexico City, Mexico
difference was found between the two methods performed (p \ 0.0025). Variance analysis also was applied to the mean values of the scores achieved by both groups (p \ 0.0001). Conclusions A new laparoscopic training method has been developed. It applies real-time dynamic optics that trainees assist by means of a mechatronic device harnessed to their body. This new training tool allows resident trainees to adapt quickly to the work environment of dynamic optics and thus enter the surgical scenario more rapidly and confidently after mastering the visual-spatial aspect of the laparoscopic approach. Keywords Dynamic perspective Laparoscopic simulator Laparoscopic training
Laparoscopic surgery demands of surgeons special skills acquired only through practice. Laparoscopic learning and training systems traditionally have an optical system that, once positioned, remains fixed and cannot refresh the perspective unless the task is interrupted and the camera manually repositioned. During surgery, the visual perspective changes constantly to relocate the surgical target. The visual perspective is dynamic. This slight difference is a limitation for any novice surgeon. It is important, therefore, that surgeons adapt, early in their training, to a changing visual perspective because it is part of the natural work environment in the operating room. This report proposes the use of a mechatronic system that allows the trainee to handle optics dynamically during training in the cutting task and thus adapt to dynamic relocation of the surgical target. The study was conducted in two phases. The first phase involved using fixed optics to cut a circle drawn on a piece of cloth. The second phase
123
1342
involved the same cutting task but with the visual perspective changed dynamically by the user via a mechatronic assistant. The data show that by adapting to dynamic optics, medical trainees can quickly and easily handle and locate the task with real-time changes in visual perspective, and also can improve task quality. A significant statistical difference was found between the two methods performed (p \ 0.0025). Variance analysis also was applied to the mean values of the scores achieved by both groups (p \ 0.0001). In conclusion, a new laparoscopic training method has been developed. It applies real-time dynamic optics that trainees assist by means of a mechatronic device harnessed to their body. This new training tool allows resident trainees to adapt quickly to the work environment of dynamic optics and thus enter the surgical scenario more rapidly and confidently after mastering the visual-spatial aspect of the laparoscopic approach. In the past few decades, laparoscopic surgery has confirmed both its importance and its advantages over open surgery [1, 2]. It offers great benefits but requires in turn that specialty surgeons maintain and refine the technique throughout their professional career [3, 4]. Studies have shown that laparoscopic skills are developed through practice, which usually is acquired on trainers [5] available in three basic types. Systems of the first type are based on a computer-generated virtual reality [6–8] that allows for sensory interaction [9] and provides trainees real-life surgical abilities without subjecting them to the stress of traumatizing a patient. These systems intrinsically offer evaluation methods in the form of games, simple tasks, or complete virtual surgeries with tactile realism. Systems of the second type are physical trainers [10, 11] that comprise mini-cameras and semicylindrical spaces. Such systems allow for training on animals or inanimate synthetic models [12] and are inexpensive and versatile. Some of these systems offer manual control of 08 and 308 optics [13]. Others are simply a mirrored box setup [14]. The third type of training system involves a hybrid approach that combines real and virtual feedback. Such systems allow trainees to use real instruments on a totally simulated model. These new-generation trainers combine the benefits of box trainers and virtual-reality trainers [15]. All trainers and simulators currently on the market, be they the physical, virtual-reality, or hybrid type [5], have certain advantages and disadvantages. We believe, nevertheless, that they all have a common disadvantage: The training takes place using a fixed visual perspective. This can limit some still unconsidered learning and training objectives. All these traditional training systems use a fixed visual perspective. That is, once the target is visualized through the optics, the optics remain static during
123
Surg Endosc (2010) 24:1341–1346
execution of the entire task. In traditional laparoscopic surgery, however, the visual perspective changes; it is dynamic. The surgeon needs constantly to reposition the optical perspective with respect to the surgical target and instruments. This difference is an initial limitation for novices, whether they are surgical assistants or surgeons, because a constantly changing perspective causes them to confuse the mental scenario. This report proposes a new training system that uses dynamic optics. It describes the experience of using a physical trainer together with a computer system that evaluates the cutting task and a mechatronic assistant [16] that allows users to assist themselves in changing the optical perspective in real time. This report shows that by controlling the laparoscope while performing a task, trainees generate a better mental scenario of the work space and improve placement of the laparoscopic instruments.
Materials and methods Adaptation to a dynamic visual perspective was analyzed using a physical trainer with several entry ports designed at the Centro de Investigacion y de Estudios Avanzados (CINVESTAV) of the Instituto Politecnico Nacional (IPN) [17]. This laparoscopic surgery trainer was built based on the shape of the abdominal cavity formed during surgery. The visual feedback is achieved using a color mini-camera with 480-line resolution, 08 optics, and a 20-in. commercially available television. A Postural Mechatronic Assistant for Solo Surgery [18], also designed at the CINVESTAV of the IPN, was used so the trainees themselves could dynamically change the optical perspective (Fig. 1). This cross-design study involved 26 medical surgeons enrolled in a laparoscopy course at the Hospital Los Angeles in Mexico City, who, without any prior experience in laparoscopic surgery, performed the cutting task 30 times each. The study was performed over a 4-month period, which was the duration of the laparoscopy course. During the first 2 months, the doctors used fixed optics. Throughout the remaining 2 months, they used dynamic optics. Exercises involving repetition were performed at 5min intervals if the availability of the doctors allowed for it. All the trainees were right-hand dominant. The doctors enrolled in the course were given a lecture on the techniques of laparoscopic surgery before joining the protocol. The lecture covered current procedures and included a physical demonstration of how the cutting task is performed in the trainer. The mechatronic assistant consists of a harness placed on the chest of the surgeon and a mechanical arm connected to the harness. The arm, driven by a DC motor, is activated by floor pedals to make up and down changes in
Surg Endosc (2010) 24:1341–1346
1343
Fig. 1 Movements. A Right and left. B In and out. C Up and down
the optical perspective. The pedals move the laparoscope, which is attached to the end of the arm, thus modifying the angle of view at the point of entry. Likewise, the postural movements of the surgeon cause the laparoscope to move in and out of the work area, and the motion of the torso shifts the view left and right. The chosen task was the cutting task proposed by McGill University [19] because it requires visual-spatial mastery and precision as well as use of the nondominant hand. A circle with a diameter of 4 cm was cut from a piece of gauze measuring 10 9 10 cm. The gauze was held by alligator clips, and the circle was cut using curved scissors and generic graspers. Homemade computer software based on Matlab (The MathWorks Headquarters, Natick, MA) was used to evaluate task time and precision, and also to perform the statistical analysis of the data. The time score was calculated by subtracting the time required to complete the exercise from a preset cutoff time: time score = cutoff time (s) minus the time required to complete the exercise (s). The precision score was calculated by penalizing the percentage of deviation from the area of a perfect circle compared with the circle cut from the cloth, considering that both circles had the same diameter. The evaluation and analysis software used the image processing tools provided by Matlab. The study was carried out in two sequential phases. The first phase was achieved using fixed optics. The optical system was positioned and then left unmoved throughout the entire task (Fig. 2). The 26 medical trainees performed the aforementioned cutting task in random order. The computer assessed task time and precision by calculating the percentage of area deviation in cutting the cloth circle. The second phase involved the same 26 medical residents, who next performed the same cutting task but with a
Fig. 2 Cutting task using fixed optics
visual perspective that they changed dynamically using the mechatronic assistant (Fig. 3). The residents strapped the mechatronic assistant to their bodies using a harness and handled the camera posturally to visually locate the work space. They performed the task seeking the best visual perspective of the surgical target that the mechatronic system could possibly offer (Fig. 4). They combined the basic visual approach movements with those of the mechatronic assistant: right, left, up, down, in, and out. The measured times for each of the 31 repetitions performed by the 26 trainees in the two phases were used to calculate the average time taken for each repetition. The average precision score was calculated likewise. (Figs. 5, 6). The trial conditions, equipment, and configurations were the same for all the participants. There were no changes in the image brightness, resolution, or quality with
123
1344
Surg Endosc (2010) 24:1341–1346
Fig. 5 Mean time and standard deviation for evolution, adaptation to, and mastery of the technique with a fixed (A) or dynamic (B) visual perspective
the video monitor. The camera’s angle of vision was the same for all fixed-vision cases. The camera was centered on the objective at the same height, positioned there by an expert surgeon.
Results
Fig. 3 Changing the visual perspective with the mechatronic assistant
Fig. 4 Cutting task. Use of the mechatronic assistant to obtain a different visual approach to the cutting area
123
The first group of residents was able to perform the cutting task uneventfully without using the mechatronic assistant. The results of the first phase show quick adaptation to instrument location in a two-dimensional space. It is
Surg Endosc (2010) 24:1341–1346
Fig. 6 Mean precision scores for the trainees who performed the task using fixed (A) or dynamic (B) optics
presumed that after the eighth repetition, the residents have mastered the technique and achieved visual-spatial adaptation to the work space. After the 25th repetition, they have reduced task execution time by approximately 40%. In this study, the precision score was low during the first six repetitions but rose by up to 25 points after the sixth repetition. The results agree with the number of tasks needed to master the technique [20–22]. In the second phase, the residents used the mechatronic assistant. As shown in Fig. 5, only the first three repetitions indicate a lack of adaptation and increased task performance time. The results exhibit a progressive decrease in the time taken to perform the task and a progressive increase in task quality (precision score) until both parameters steady at mean values (Fig. 6). The performance curves (time vs. iteration number) show that for sets of 30 iterations, the medical trainees reached 90% of the plateau value at iteration number 11 with fixed optics and at iteration number 5 with dynamic optics. The scores show that the trainees reached 90% of the plateau value at iteration number 9 with fixed optics and at iteration number 5 with dynamic optics. The differences between the mean values were examined for statistical significance using two-factor variance analysis, in which the first factor was the fixed optics of the laparoscope and the second factor was the dynamic optics of the mechatronic assistant. The results show a significant statistical difference between the two methods (p \ 0.0025). Variance analysis also was applied to the mean values of the scores achieved by both groups (p \ 0.0001). Significant differences between the two methods were identified.
Discussion A new laparoscopic training method was tested with 26 residents in their final year of the specialty. The method calls for trainees to perform the cutting task on a physical
1345
trainer, adjusting the optics themselves with the help of a mechatronic system. The same task was performed for reference applying normal training conditions and fixed optics. In both tests, 08 optics was used. The results obtained with the fixed optics were similar to those reported in the literature and thus provided a suitable reference for this study. The results of trainee visual selfassistance show that a time delay occurs initially in all cases. We believe this is due to a lack of motor skills in using the mechatronic system to reach the work space and a visual-spatial adaptation to establish the work space and then optimize the approach to the target. The time results show that once the motor and visual-spatial skills are mastered, the average time required to perform the task stabilizes at 84 s, and the precision score, which initially is low, rises to a maximum value of 87, which is higher than the maximum score of 82 obtained with fixed optics. A normal surgical scenario has several visual stages: an initial complete exploration; assistance in the placement of ports; constant zooming in and out to reach the surgical targets; and finally, another complete exploration. This project found that when the residents themselves assisted the optics, they were able to zoom out of the surgical scenario consistently and to zoom progressively into the target with a natural mastery of the motor and visual-spatial skills and without detriment to the quality of the task. Currently, no system is available on the market that offers real-time optics readjustment for residents to use in training to zoom in and out of the work space and surgical target, so it was not possible to establish the importance of dynamic handling of optics during learning and training. We believe that a trainee-assisted visual system for learning and training will result in faster adaptation to the surgical scenario and greater precision in the task, as shown in this study.
Conclusion A new training method has been developed using real-time dynamic optics. The system offers a computer software objective evaluation of task time and a precision score for the laparoscopic cutting task. The trainee locates the visual perspective of the work area intuitively by means of slight postural movements and without having to stop the task, thus achieving a better visual approach to the surgical target. The results obtained show that by adapting to dynamic optics, residents are able to handle and locate the task with real-time changes in optical perspective, and also to improve task precision. We believe this new training tool will allow residents to enter the surgical scenario as surgical assistants or surgeons faster and more confidently after they have mastered the visual-spatial aspect of the
123
1346
laparoscopic approach. The estimated cost of the mechatronic system used in this study is less than US$1,000. Disclosures Arturo Minor Martı´nez, Jose´ Luis Limo´n Aguilar, Ricardo Ordorica Flores, Jose´ Luis Ortiz Simo´n, and Alejandro Garcı´a Pe´rez have no conflicts of interest or financial ties to disclose.
References 1. Bergamaschi R, Marvik R, Johnsen G, Thoresen JE, Ystgaard B, Myrvold HE (1999) Open vs laparoscopic repair of perforated peptic ulcer. Surg Endosc 13:679–682 2. Pokata N, Delaney CP, Senagore AJ, Brady KM, Fazio VW (2005) Laparoscopic vs open total colectomy. Surg Endosc 19:531–535 3. Koh CH, Janik GM (1999) Laparoscopic microsurgery: current and future status. Curr Opin Obstet Gynecol 11:401–407 4. Gutt CN, Kim ZG, Kra¨henbu¨hl L (2002) Training for advanced laparoscopic surgery. Eur J Surg 168:172–177 5. Undre S, Darzi A (2007) Laparoscopy simulators. J Endourol 21:274–279 6. Hamilton EC, Scott DJ, Fleming JB, Rege RV, Laycock R, Bergen PC, Tesfay ST, Jones DB (2002) Comparison of video trainer and virtual reality training systems on acquisition of laparoscopic skills. Surg Endosc 16:406–411 7. Torkington J, Smith SGT, Rees BI, Darzi A (2001) Skill transfer from virtual reality to a real laparoscopic task. Surg Endosc 15:1076–1079 8. Hochberger J, Euler K, Naegel A, Hahn EG, Maiss J (2004) The compact Erlangen active simulator for interventional endoscopy: a prospective comparison in structured team-training courses on ‘‘endoscopic hemostasis’’ for doctors and nurses to the ‘‘EndoTrainer’’ model. Scand J Gastroenterol 39:895–902 9. Coleman J, Nduka CC, Darzi A (1994) Virtual reality and laparoscopic surgery. Br J Surg 81:1709–1711
123
Surg Endosc (2010) 24:1341–1346 10. Minor A, Lorias D (2007) Novel laparoscopic home trainer. Surg Laparosc Endosc Percutan Tech 17:300–302 11. Pokorny MR, MClaren SL (2004) Inexpensive home-made laparoscopic trainer and camera. ANZ J Surg 74:691–693 12. Chung JY, Sackier JM (1998) A method of objectively evaluating improvements in laparoscopy skills. Surg Endosc 12:1111–1116 13. Minor MA, Ordorica FR, Ortiz SJ (2009) Thirty-degree optical system for laparoscopic training. Minim Invasive Ther Allied Technol 18(4):1365–2931 14. Keyser EJ, Derossis AM, Antoniuk M, Sigman HH, Fried GM (2000) A simplified simulator for the training and evaluation of laparoscopic skills. Surg Endosc 14:149–153 15. Haptica (2009) The world-leading ProMIS surgical simulator. Retrieved 12 November 2008 at http://www.haptica.com/ index.htm 16. Minor A, Mun˜oz R, Nieto J, Ordorica R (2005) Postural mechatronic assistant for laparoscopic training. Minim Invasive Ther Allied Technol 14:357–359 17. Martinez AM, Espinoza DL (2007) Novel laparoscopic home trainer. Surg Laparosc Endosc Percutan Techn 17:300–302 18. Minor A, Ordorica R, Villalobos J, Galan M (2009) Device to provide intuitive assistance in laparoscopic holding. Ann Biomed Eng 37:643–649 19. Derossis AM, Fried GM, Abrahamowicz M, Sigman HH, Barkun JS, Meakins JL (1998) Development of a model for training and evaluation of laparoscopic skills. Am J Surg 175:482–487 20. Brunner W, Korndorffer J Jr, Sierra R, Massarweh N, Dunne J, Yau C, Scott D (2004) Laparoscopic virtual reality training: are 30 repetitions enough? J Surg Res 122:150 21. Grantcharov TP, Bardram L, Funch-Jensen P, Rosenberg J (2003) Learning curves and impact of previous operative experience on performance on a virtual reality simulator to test laparoscopic surgical skills. Am J Surg 185:146–149 22. Madan AK, Harper JL, Taddeucci RJ, Tichansky DS (2008) Goal-directed laparoscopic training leads to better laparoscopic skill acquisition. Surgery 144:345–350
