Quantitative Comparison of Surgical Device Usage in Laparoscopic Gastrectomy Between Surgeons' Skill Levels: an Automated Analysis Using a Neural Network.

Background: Whether surgical device usage in laparoscopic gastrectomy differs with respect to operator's skill levels is unknown. Further, device usage analysis using artificial intelligence has not been reported to date. Herein, we compared the patterns of surgical device usage during laparoscopic gastrectomy for gastric cancer among surgeons at different skill levels. The data of device usage was acquired from laparoscopic video recordings using an automated surgical-instrument detection system.

Methods: In total, 100 video recordings of infrapyloric lymphadenectomy and 33 of D2 suprapancreatic lymphadenectomy during laparoscopic gastrectomy for gastric cancer were analyzed in this retrospective study. The system's accuracy was evaluated by comparing the automatic and the manual usage time. Surgical device usage patterns were compared between qualified and nonqualified surgeons of The Japan Society for Endoscopic Surgery Endoscopic Surgical Skill Qualification System.

Results: For every device, the automatic detection time and manual detection time were consistent with each other. In infrapyloric lymphadenectomy, the usage time proportions of dissector forceps and clip applier were higher among nonqualified operators than among qualified operators (dissector, 5.1% vs. 2.3%, P < 0.001; clip applier, 1.6% vs. 1.3%, P < 0.01). In suprapancreatic lymphadenectomy, the usage time proportions of energy devices, clip applier, and grasper forceps were significantly different (energy devices, 59.6% vs. 50.6%, P < 0.001; clip applier, 1.4% vs. 0.9%, P < 0.001; only grasper forceps; 18.3% vs. 27.9%, P = 0.022).

Conclusions: Quantitative analysis of laparoscopic device usage using the automated surgical device detection system showed that the patterns of device usage during laparoscopic gastrectomy differed depending on surgeons' skill levels. These differences could suggest how the qualified and nonqualified surgeons performed the procedures.