Virtual Reality Warm-up Improves Robotic Surgery Performance and Reduces Errors - Medical Training Magazine

Virtual Reality Warm-up Improves Robotic Surgery Performance and Reduces Errors

Thomas S. Lendvay, MD, FACS reports on how virtual reality, robotic pre-surgical rehearsal improves performance and reduces errors in both surgical trainees and experienced surgeons.

Demonstration of our set-up in the ISIS simulation center with the da Vinci. Tasks are affixed to acrylic plates so that the orientation of the robot remains constant between modules. (Resident: Daniel Avery, Urology PGY-4)
Demonstration of our set-up in the ISIS simulation center with the da Vinci. Tasks are affixed to acrylic plates so that the orientation of the robot remains constant between modules. (Resident: Daniel Avery, Urology PGY-4)

Efforts to reduce the more than 98,000 US lives lost annually from medical complications have included the use of simulation curricula for teaching novice and practicing clinicians basic and advanced skills. [1, 2] Access to surgical simulation for residents has been mandated by the American Board of Surgery among other professional societies [3, 4]. Presurgical rehearsal or warm-up has been shown to enhance surgical performance and reduce operative times. Now that high fidelity simulators have been created for robotic surgery, we sought to explore whether virtual reality (VR) robotic surgical warm-up would improve performance and reduce errors in both surgical trainees and experienced surgeons.

Athletes and musicians warm-up, so why don’t surgeons? Surgery is a high stakes, technically challenging, and cognitively intense profession. In addition, the benefits of warm-up may be particularly important for robotic surgery due to the increased information presented to the surgeon through the visual monitor. Surgeons need to process and transform visual cues into forces applied because of the lack of haptic feedback from the platform.

METHODS

Surgical residents and faculty from the Departments of Urology, General Surgery, and Gynecology at the University of Washington Medical Center (UWMC) and Madigan Army Medical Center (MAMC) were recruited for our study. [Table 1.] PGY-1 and 2 residents were excluded as we felt that robotic surgery performance was not level appropriate for junior residents. We enrolled 51 subjects and after a proficiency curriculum to bring every subject to a baseline of VR and da Vinci (Intuitive Surgical Inc., Sunnyvale, California) robotic surgical skills level, we randomized them to a series of trial sessions either exposing or not exposing them to a brief VR robotic simulator (dV-Trainer simulator, MIMIC Technologies, Inc., Seattle, Washington.) warm-up.

MIMIC dV-Trainer VR simulation mode\\ules from left to right: Pick and Place, Ring Walk Level 1, Pegboard Level 1, Pegboard Level 3.
MIMIC dV-Trainer VR simulation mode\\ules from left to right: Pick and Place, Ring Walk Level 1, Pegboard Level 1, Pegboard Level 3.

The proficiency curriculum included four VR and four da Vinci dry lab modules that exercised basic robotic manipulations such as instrument and camera clutching, suturing, object transfer between instruments, and spatial relations capabilities. [Figures 1. and 2.] To achieve proficiency, each subject had to pass performance benchmark criteria with zero error rates to advance to the next degree of task difficulty.

 Table 1. Subject demographics between control and warm-up groups. P-values denote similarities between the cohorts.

Variable Control (N=25) Warm Up (N=26) p-value
Age 35.32±6.47 33.85±5.82 0.396
Gender
   Female 10 (40.0%) 9 (34.6%) 0.691
   Male 15 (60.0%) 17 (65.4%)
Musical instrument for >3yrs
   No 7 (28.0%) 9 (34.6%) 0.611
   Yes 18 (72.0%) 17 (65.4%)
Handedness
   Ambidextrous 0 (0.0%) 1 (3.8%) 0.216
   Left 2 (8.0%) 0 (0.0%)
   Right 23 (92.0%) 25 (96.2%)
Training Year
   PGY1 1 (4.0%) 0 (0.0%) 0.299
   PGY2 0 (0.0%) 2 (7.7%)
   PGY3 4 (16.0%) 9 (34.6%)
   PGY4 3 (12.0%) 1 (3.8%)
   PGY5 3 (12.0%) 1 (3.8%)
   PGY6 2 (8.0%) 1 (3.8%)
   Faculty 12 (48.0%) 12 (46.2%)
Sub Specialty
   Urology 14 (56.0%) 14 (53.8%) 0.568
   General Surgery 7 (28.0%) 5 (19.2%)
   OBGYN 4 (16.0%) 7 (26.9%)
Recent Video Game Use
   None 15 (60.0%) 16 (61.5%) 0.890
   <2 x Week 7 (28.0%) 6 (23.1%)
   2+ x Week 3 (12.0%) 4 (15.4%)
Laparoscopic Cases (primary surgeon)
   None 1 (4.0%) 0 (0.0%) 0.497
   10 or less 3 (12.0%) 3 (11.5%)
   11-25 3 (12.0%) 1 (3.8%)
   25+ 18 (72.0%) 22 (84.6%)
Robotic Cases (primary surgeon)
   None 9 (36.0%) 8 (30.8%) 0.564
   10 or less 6 (24.0%) 10 (38.5%)
   11-25 3 (12.0%) 1 (3.8%)
   25+ 7 (28.0%) 7 (26.9%)

We designed a da Vinci tool tracking method – SurgTrak™ – to locate the tools in space so we could get path length data to calculate economy of motion. [Figure 3.] Using this technology, we had already demonstrated construct validation of the proficiency curriculum [1].

Four trial sessions per subject were performed. The first three tested a similar VR to dry lab task – the rocking peg board. And the fourth session tested a dissimilar task from the rocking peg board warm-up – da Vinci suturing. [Figure2.] The warm-up took 3-5 minutes to complete and the control subjects spent 10 minutes reading a leisure book immediately prior to performing the da Vinci criterion task. We tracked total task time, path length for right and left handed tools, technical errors, cognitive errors, and economy of motion for each of the sessions on the simulator and on the da Vinci.

RESULTS

The warm-up group performed with decreased task time (-29.29 seconds, p=0.001) and path length (-79.87 mm, p=0.014) for the similar tasks. There was a >6-fold reduction favoring warm-up for sessions with errors of placing the rings on incorrect pegs (sequence errors). In Tables 2 and 3, performance metrics for the first three similar sessions are detailed.

image2
Figure 2. da Vinci dry lab modules from left to right: Fundamentals of Laparoscopic Surgery (FLS) block transfer, FLS intracorporeal suturing (this was the criterion task for sessions #4), Ring tower (The Chamberlain Group, Great Barrington, Massachusetts), Rotating rocking pegboard (this was the criterion task for sessions #1-3).

 

 

 

 

image3
Figures 3. Retrofitted da Vinci training instrument with electromagnetic sensor housing on back end (left) and task module jig that housed modules and transmitter in fixed distances from each other for standard tool tracking (right).

Table 2. Outcomes by study group for sessions 1-3, analyzed with repeated measures ANOVA.

Control Warm-Up
Outcome Mean SE Mean SE Difference (95% CI) P-Value
Economy of Motion 4.42 0.1 4.63 0.1 0.21 (-0.06, 0.47) 0.132
Task Time 264.31 6.49 235.01 6.36 -29.29 (-47.03, -11.56) 0.001
Total Peg Touches 21.68 1.63 19.38 1.59 -2.29 (-6.71, 2.12) 0.313
Cognitive Error 0.12 0.04 0.06 0.04 -0.06 (-0.17, 0.06) 0.340
Path Length 1149.23 23.27 1069.37 22.71 -79.87 (-144.48, -15.25) 0.014

Table 3. Binary outcomes for sessions 1-3 by study group and relative risk (RR) of errors by type, analyzed with relative risk regression.

Proportion of Sessions with Error
Error Type Control Warm-Up RR 95% CI P-Value
Ring Drops 0.320 0.333 0.96 (0.58, 1.59) 0.873
Air Transfer 0.040 0.051 0.78 (0.19, 3.14) 0.727
Out of Order (Sequence) 0.080 0.013 6.24 (0.77, 50.76) 0.087

When we tested whether the dissimilar VR task can warm-up subjects for robotic suturing, we observed a 4-fold reduction in the proportion of sessions with global technical errors for the suturing (needle entrance, exit errors and air knot errors, collectively, p=0.020). [Table 4.]

Table 4. Outcomes by study group for session 4, analyzed with t-tests. Global Technical Error = composite of Air knot, Needle targeting errors by FLS (Entrance and Exit dots errors).

 

Control (n=25) Warm Up (n=26)
Outcome Mean SD Mean SD P-Value
Task Time 111.20 5.85 107.58 7.41 0.703
Economy of Motion 3.69 0.17 3.82 0.16 0.560
Path Length 401.42 22.89 401.46 26.97 0.999
Global Technical Error 0.44 0.12 0.12 0.06 0.020

 

When we divided the groups by level of MIS experience (> 10 laparoscopic and > 10 robotic cases as primary surgeon vs. < 10 cases in each modality experience), we observed that the warm-up effect was more pronounced with experience. Economy of motion (p=0.007), task time (p=0.001), and path length (p=0.093) favored the warm-up ‘experienced’ sub-group. [Table 5.]

Table 5. The table below shows the mean values and differences (Warm-Up – Control), 95% confidence intervals, and p-values for the 5 continuous outcomes measured in sessions 1-3 in the study overall and broken up by Robotic/Laparoscopic Case experience.

 

<= 10 Rob & Lap (N=34) > 10 Robotic and Laparoscopic Cases (N=17)
Outcome Warm up Control Difference (95% CI) P Warm up Control Difference (95% CI) P
Economy of Motion 4.51 (0.11) 4.49 (0.12) 0.02 (-0.3, 0.34) 0.897 4.94 (0.18) 4.31 (0.15) 0.63 (0.18, 1.09) 0.007
Task Time 240.8 (7.2) 258.6 (8.1) -17.8 (-39.2, 3.5) 0.102 219.4 (11.9) 272.9 (10.0) -53.5 (-83.9, -23.0) 0.001
Peg Touches 20.7 (1.8) 24.2 (2.0) -3.6 (-8.8, 1.7) 0.184 15.9 (2.9) 17.9 (2.4) -2 (-9.4, 5.5) 0.600
Cognitive Errors 0.05 (0.05) 0.13 (0.05) -0.08 (-0.22, 0.06) 0.265 0.10 (0.08) 0.10 (0.07) 0 (-0.21, 0.20) 0.963
Path Length 1077 (27) 1152 (30) -75 (-154, 4) 0.063 1049 (44) 1145 (37) -97 (-210, 16) 0.093

DISCUSSION and CONCLUSIONS

Warming-up on a VR platform before doing similar and dissimilar skills modules on the da Vinci improved surgeon performance. It is intuitive that some form of priming before doing a task would be beneficial, but the finding that dissimilar tasks can elevate surgeon performance is compelling to bring this into the operating room. Testing whether warm-up improved robotic suturing was critical because the suturing task was analogous to suturing that would be performed in human surgeries. The value of this finding is that the ideal warm-up curricula may not need to look like the planned robotic surgery tasks – generalizability. Furthermore, the observation that experienced surgeons derived more of a benefit than inexperienced surgeons, makes this finding relevant to the practicing clinician and not just the trainee. This finding parallels Mucksavage et al.’s finding that laparoscopic warm-up decreases operative times in experienced surgeons in the OR [2],

Our observation that warm-up improved not only technical, but trended towards improving cognitive skill as well suggests that recruiting not only simple psychomotor centers of the brain, but also spatial relations centers, may be additive to the warm-up benefit.

Our study design yielded some limitations. Because of the variability in skill level between different subjects despite passing the proficiency curriculum, we recognize that if the better performing surgeons tended to be randomized to the warm-up group, this could have skewed the findings. One way to avoid this confounder in the future is to design the study with each subject as their own controls either receiving of not receiving the warm-up through a series of sessions. The only challenge here would be, however, that the sequence of either receiving or not receiving warm-up would have to be varied so that prior sessions do not yield a practice effect. In addition, intervals between sessions varied among subjects because of the difficulty in scheduling times for active residents and faculty. This could have led to performance variability as Jenison et al. showed that after 4 weeks of rest, robotic surgery skills degrade, thus we minimized the number of intervals that exceeded this threshold. [3] We did not also control for fatigue as the subject’s post-call level was not questioned or recorded.

Our findings were unambiguous in a dry lab setting, yet demonstration in the operating room as Mucksavage et al. and Calatayud et al. did for conventional laparoscopy, needs to be performed. [2,4] Because the desktop dV-Trainer runs the same software as the Intuitive backpack simulator, our findings may be easily translatable into the OR. In future we envision a generation of robotic systems capable of downloading patient specific images (from CT or MRI scans) to allow a surgeon to sit at the console and rehearse the case through performance on a 3-D VR rendered anatomy module. [5]

Robotic VR simulation warm-up improves technical and cognitive performance on the da Vinci robot in a dry lab setting. These results provide a foundation for predictive validation studies in the OR evaluating the role of warm-up for improving surgical outcomes, reducing operative cost, and paving the way for patient-specific procedure rehearsal

ACKNOWLEDGEMENTS

This study was supported by the Department of Defense U.S. Army Medical Research and Materiel Command under award number W81XWH-09-1-0714 (PI: Lendvay). Views and opinions of, and endorsement by the author(s) do not reflect those of the Army or the Department of Defense. The Seattle Children’s Core for Biomedical Statistics is supported by the Center for Clinical and Translational Research at Seattle Children’s Research Institute and grant UL1RR025014 from the NIH National Center for Research Resources.

About the Author:

Thomas S. Lendvay, MD, FACS, is an associate professor of urology, University of Washington School of Medicine, Department of Urology. He may be contacted at thomas.lendvay@ seattlechildrens.org.

 REFERENCES

  1. Tausch TJ, Kowalewski TM, White LW, McDonough PS, Brand TC, Lendvay TS. Content and construct validation of a robotic surgery curriculum using an electromagnetic instrument tracker. J Urol 188:919-923, 2012.
  2. Muchsavage P, Lee J, Kerbl D, Clayman R, McDougall E. Preoperative warming up exercises improve laparoscopic operative times in an experienced laparoscopic surgeon. J Endourol 26:765-768, 2012.

3 Jenison EL, Gil KM, Lendvay TS, Guy MS. Robotic surgical skills: Acquisition, maintenance, and degradation. JSLS 16:218-228, 2012.

  1. Calatayud D, Arora S, Aggarwal R, et al.Warm-up in a virtual reality environment improves performance in the operating room. Ann Surg. 251:1181, 2010.
  2. Makiyama K, Nagasaka M, Inuiya T, et al. Development of a patient-specific simulator for laparoscopic renal surgery. Int J Urol 19:829-835, 2012.