A clinically and technologically practicable navigation system for robotic sacrocolpopexy.

Abstract

<h4>Purpose</h4>Pelvic organ prolapse is characterized by the descent of one or more pelvic organs toward the vaginal canal. Surgical correction through sacrocolpopexy requires precise mesh anchoring to the sacral promontory, where vessels hidden by adipose tissue, but rigidly fixed to the bone, pose a major risk. This study introduces an augmented reality (AR) navigation system for vessel visualization at the anchoring site, integrated into the Da Vinci robotic console and avoiding intraoperative imaging.<h4>Methods</h4>Anatomical landmarks visible in both preoperative CT scans and intraoperative endoscopic views were identified as the insertions of the round ligaments into the inguinal canal. Their 3D positions were determined by stereo endoscopic triangulation. As two landmarks are insufficient for full rigid registration, a dedicated rod connecting the endoscope to the pubic symphysis was designed to provide the missing degree of freedom. Our algorithm aligns the anatomical model with the endoscopic coordinate system by iteratively minimizing the distance between the virtual symphysis and the rod while preventing interpenetration.<h4>Results</h4>In a phantom-based feasibility study, the method enables geometric alignment of preoperative CT-based vascular models with intraoperative endoscopic views with an average error on the order of 10 mm. This accuracy falls within the clinically safe vascular window at the promontory, ensuring consistent overlay during mesh placement, with the rod serving as a stable and robot-compatible reference.<h4>Conclusion</h4>The presented system enables AR visualization of vascular anatomy during robotic sacrocolpopexy, providing anatomical guidance through a fully minimally invasive workflow. This represents the first attempt to apply AR-based navigation in gynecological surgery. While this study demonstrates the technical feasibility of the approach, future work will focus on validation across multiple anatomical models and in vivo clinical settings.