Patients-Specific Virtual Surgical Navigation for Lung Segmentectomy: A Prospective Multicenter Study

Abstract

Background: Anatomical segmentectomy has become a standard surgical option for small peripheral non-small cell lung cancer (NSCLC); however, its technical complexity necessitates precise preoperative planning. This study evaluated the feasibility and clinical utility of the lung module of a patient-specific virtual surgical navigation system for preoperative planning in anatomical segmentectomy and subsegmentectomy. Methods: This prospective multicenter observational study enrolled 34 patients undergoing anatomical segmentectomy or subsegmentectomy between May and July 2025. Preoperative planning was sequentially performed using conventional two-dimensional (2D) CT and a patient-specific virtual surgical navigation system based on AI-driven three-dimensional (3D) reconstruction. Feasibility outcomes included turnaround time, operational stability, and accuracy of tumor localization, segment prediction, and bronchovascular anatomy compared with intraoperative findings. Surgeon workload was assessed using the NASA Task Load Index (NASA-TLX). Perioperative outcomes were compared with historical cohorts planned using conventional 2D CT and commercially available 3D CT systems after propensity score matching. Results: All cases achieved successful 3D reconstruction within 72 hours, with complete operational stability. The navigation system demonstrated near-perfect concordance with operative findings for tumor localization and segment prediction (κ = 0.96–1.00), and significantly higher accuracy in predicting resected arteries, veins, and bronchi compared with 2D CT planning. Surgeon workload was significantly reduced with navigation system–based planning (overall NASA-TLX score: 52.5 ± 12.1 vs. 76.1 ± 15.1; p < 0.001), particularly in mental, physical, and temporal demand domains. Compared with both 2D CT and conventional 3D CT planning cohorts, use of the navigation system was associated with shorter operative time, reduced blood loss, and fewer resected subsegments, while maintaining comparable surgical margins and postoperative outcomes. Conclusion: The lung module of a patient-specific virtual surgical navigation system is a feasible and effective tool for preoperative planning in anatomical segmentectomy and subsegmentectomy. It improves anatomical prediction accuracy, reduces surgeon workload, and demonstrates favorable perioperative performance, supporting its clinical value in technically demanding lung-sparing surgery for early-stage NSCLC.