Thoracoscopic surgery for congenital pulmonary malformations in infants: a preliminary comparative evaluation of virtual reality-guided preoperative planning versus standard imaging.

BACKGROUND: Thoracoscopic resection of congenital pulmonary malformations remains technically demanding because of limited working space, variable broncho-vascular anatomy and the need to preserve healthy lung parenchyma. An integrated workflow combining patient-specific three-dimensional (3D) segmentation with immersive virtual reality (VR) navigation may enhance preoperative planning, but clinical data remain limited. METHODS: We performed a retrospective historical cohort study of children undergoing thoracoscopic resection between 2013 and 2024. Patients were grouped according to planning strategy: standard imaging (SI) alone or an additional integrated workflow combining patient-specific 3D segmentation with immersive VR navigation (VR). Perioperative outcomes were compared, with focused analyses of lesion and procedure heterogeneity and chronological effects. RESULTS: Fifty infants were analyzed (39 SI, 11 VR). Median operative time was 183 versus 248 min (p = 0.42). Conversion occurred in 10/39 SI cases and no VR cases (p = 0.045); intraoperative complications occurred in 5/39 and 0/11 (p = 0.20), and postoperative complications in 6/39 and 0/11 (p = 0.16). Median NICU/PICU stay was 2 versus 1 day (p = 0.045), and hospital stay 8 versus 5 days (p < 0.001). Three SI patients required redo surgery and no VR patients did. Findings showed substantial case-mix heterogeneity, but their favorable direction was maintained in focused analyses. VR cases occurred later in the institutional experience. CONCLUSIONS: The integrated VR-guided workflow was feasible and associated with favorable perioperative signals. Although the historical design precludes causal attribution, the consistency of these preliminary findings supports prospective evaluation of VR as a patient-specific planning adjunct.