Initial experience with the EndoFusion 3D system for noninvasive intraoperative localization of pulmonary nodules.

BACKGROUND: Intraoperative localization of small peripheral pulmonary nodules remains difficult during video-assisted thoracoscopic surgery (VATS), especially for subcentimeter and ground-glass-predominant lesions. Conventional computed tomography (CT) guided localization is effective but invasive and carries radiation exposure and puncture-related risks. This study assessed the feasibility and safety of EndoFusion, a system that projects preoperative three-dimensional (3D) information onto real-time thoracoscopic views to assist localization. METHODS: A single-center retrospective study was conducted on patients undergoing VATS sublobar resection with EndoFusion-assisted localization. Preoperative 3D models were generated from thin-slice CT and registered to the pleural surface intraoperatively. Technical success, localization accuracy, perioperative outcomes, and complications were assessed. RESULTS: Fifteen patients with a total of 19 peripheral pulmonary nodules were included. The median nodule size was 8.0 mm and the median pleural depth was 3.0 mm, with most lesions characterized as pure ground-glass nodules (pGGNs) or consolidation-to-tumor ratio (CTR) ≤0.25. All patients successfully underwent VATS resection without conversion, including wedge resection (60.0%), combined wedge and segmentectomy (26.7%), or segmentectomy (13.3%). The median operative time was 75 minutes, and EndoFusion registration was completed in a median of 128 seconds. Intraoperative blood loss was minimal (median 50 mL), and no intraoperative complications occurred. Two patients (13.3%) developed minor postoperative complications, and the median hospital stay was 4 days. CONCLUSIONS: EndoFusion provides a feasible, safe, and fully noninvasive intraoperative localization method for VATS. By projecting patient-specific 3D anatomy directly onto the operative field, it enhances precision, reduces cognitive burden, and may serve as an effective alternative to conventional preoperative localization.