A head-mounted mixed reality platform for robotic flexible endoscope control in minimally invasive cholecystectomy: an animal feasibility study.

Robotic surgical systems have advanced minimally invasive surgery by enhancing dexterity and precision, yet endoscopic camera control remains a persistent bottleneck due to reliance on assistants or manual manipulation that disrupts workflow continuity. We developed MicroVision, a robotic endoscopic platform that translates natural head movements into real-time viewpoint adjustments through a head-mounted mixed reality interface. The system integrates a flexible binocular endoscope with six degrees of freedom and a position-orientation decoupled design, enabling intuitive orientation control and stable, high-definition visualization even in confined anatomical spaces. In a porcine cholecystectomy model (n = 10), MicroVision was compared with conventional laparoscopy. Verbal instructions to camera assistants were almost eliminated, dropping from 15.3 to 0.2 per procedure (P < 0.001), and mean operative time was reduced from 74.8 to 66.0 min (P = 0.038). Camera movements decreased by more than 70%, and intraoperative blood loss was reduced by nearly one quarter. All procedures were completed successfully without complications or system failures. These findings demonstrate that MicroVision enhances surgical efficiency, visualization, and workflow reliability while maintaining established safety standards, underscoring its potential contribution to robotic surgery.