RESUMEN
Robotic surgery bedside assistants play an important role in robotic procedures by performing intra-corporeal tasks while accommodating the physical presence of the robot. We hypothesized that an augmented reality headset enabling 3D intra-corporeal vision while facing the surgical field could decrease time and improve accuracy of robotic bedside tasks. Bedside assistants (one physician assistant, one medical student, three surgical trainees, and two attending surgeons) performed validated tasks within a mock abdominal cavity with a surgical robot docked. Tasks were performed with a bedside monitor providing 2D or 3D vision, or an optical see-through head-mounted augmented reality device with 2D or 3D vision. The effect of augmented reality device resolution on performance was also evaluated. For the simplest task of touching a straw, performance was generally high, regardless of mode of visualization. With more complex tasks, including stapling and pulling a ring along a path, 3D augmented reality decreased time and number of errors per task. 3D augmented reality allowed the physician assistant to perform at the level of an attending surgeon using 3D augmented reality (p = 0.08). All participants had improved times for the ring path task with better resolution (lower resolution 23 ± 11 s vs higher resolution 14 ± 4 s, p = 0.002). 3D augmented reality vision with high resolution decreased time and improved accuracy of more complex tasks, enabling a less experienced robotic surgical bedside assistant to function similar to attending surgeons. These data warrant further study with additional complex tasks and bedside assistants at various levels of training.
Asunto(s)
Realidad Aumentada , Laparoscopía , Procedimientos Quirúrgicos Robotizados , Robótica , Competencia Clínica , Humanos , Laparoscopía/métodos , Procedimientos Quirúrgicos Robotizados/métodos , Robótica/educaciónRESUMEN
INTRODUCTION: Sampling lymph nodes (LNs) is independently associated with decreased recurrence and improved survival for Wilms tumor (WT). Despite sampling, we noted cases where a few or no LNs were identified after resection of WT. We hypothesized fewer LNs were identified when submitted en bloc with the tumor, compared with when submitted separately. PATIENTS/MATERIALS AND METHODS: We conducted a retrospective chart review from 2003 to 2012 of WT resection cases, examining the type of LN dissection, the specimens submitted to pathology, number of LNs evaluated, and complications associated with the procedure. RESULTS: We identified 74 children with WT; 59 of 74 (79.7%) had unilateral disease and 15 of 74 (20.3%) had bilateral disease. With unilateral disease, more LNs were identified by separate versus en bloc sampling (5.2 ± 0.6 vs. 4.4 ± 1.2 nodes, p=0.61). Both the methods identified fewer LNs compared with en bloc+separate sampling (12.5 ± 2.7 nodes, p<0.001 and p=0.04, respectively). The majority of children with bilateral disease (10/15, 66.6%) did not have LN sampling intraoperatively. When submitted separately, 83.3 ± 3.8% of all LNs were identified in the separate specimen, and two en bloc specimens that were noted to have adenopathy intraoperatively had no LNs pathologically identified. Few cases had complications, which did not appear associated with LN sampling. CONCLUSIONS: En bloc+separate sampling yields the most LNs during resection of WT. We recommend using this technique to facilitate the maximum number of LNs evaluated in WT. Low rates of LN sampling in bilateral disease may indicate decreased regard for sampling when tumor stage is already known.