Abstract
BACKGROUND: Brain arteriovenous malformations (AVMs) present significant surgical challenges due to their complex angioarchitecture. This study evaluates the clinical application of augmented reality (AR) and 3D modeling for preoperative planning in AVM microsurgery. METHODS: A prospective series of 20 patients with Spetzler-Martin grade II-IV AVMs underwent preoperative 3D planning using computed tomography/magnetic resonance imaging segmentation (Inobitec DICOM Viewer Pro), AR navigation (Microsoft HoloLens 2 with Medgital software), and QR-code registration. Technical parameters including target registration error (TRE), preoperative preparation time, and intraoperative outcomes were analyzed. RESULTS: Complete AVM resection was achieved in all cases (100%). The AR system demonstrated submillimeter accuracy (TRE: 0.5 ± 0.3 cm with QR-code vs. 1.2 ± 0.5 cm with craniometric points). Preoperative modeling required 15.3 ± 5.1 min, while AR navigation averaged 12.7 ± 3.2 min. CONCLUSION: Our findings support the feasibility of AR-assisted 3D planning as an adjunctive tool for AVM microsurgery. This proof-of-concept study suggests that AR technology may improve spatial orientation and facilitate surgical workflow, especially for complex lesions near eloquent areas. While the clinical benefit in terms of outcomes remains to be confirmed in larger prospective studies, the presented experience demonstrates technical reliability and practical applicability. Future integration with automated segmentation and cloud-based platforms may further enhance its clinical utility.