Shray A. Patel ^1* Michael M. Covell ² Saarang Patel ³ Sai Krishna Palepu ^3* Sandeep Kandregula ^3* Avi Gajjar ^3* Oleg Shekhtman ^3* Georgios Sioutas ³ Ali Dhanaliwala ^4* Terence Gade ^4* Jan-Karl Burkhardt ^3* Visish M. Srinivasan ^3*

• ¹ Thomas Jefferson University Hospital, Jefferson University Hospitals, Philadelphia, United States
• ² School of Medicine, Georgetown University, Washington, D.C., District of Columbia, United States
• ³ Department of Neurosurgery, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, United States
• ⁴ Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, United States

Background: Extended reality (XR) includes augmented reality (AR), virtual reality (VR), and mixed reality (MR). Endovascular neurosurgery is uniquely positioned to benefit from XR due to the complexity of cerebrovascular imaging. Given the different XR modalities available, as well as unclear clinical utility and technical capabilities, we clarify opportunities and obstacles for XR in training vascular neurosurgeons. Methods: A systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines was conducted. Studies were critically appraised using ROBINS-I. Results: 19 studies were identified. 13 studies used VR, while 3 studies used MR, and 3 studies used AR. Regarding specific educational applications, VR was used for simulation in 10 studies and anatomical modeling in 3 studies. AR was only used for live intra-operative guidance (n=3 studies). MR was only used for modeling and intra-operative teaching. Considering diseasespecific uses, XR enhanced trainee understanding of intracranial aneurysms (n=12 studies) and stroke (n=7). XR trained surgeons in diverse neurosurgical procedures, including aneurysm coiling (n=5 studies), diagnostic angiography (n =5), and thrombectomy (n=5). Conclusions: Anatomical modeling with VR and MR enhances neurovascular anatomy education with patient-specific, 3-D models from imaging data. AR and MR enable live intraoperative guidance, allowing experienced surgeons to remotely instruct novices, potentially improving patient care and reducing geographic disparities. AR overlays enhance instruction by allowing the surgeon to highlight key procedural aspects during training. Inaccurate tracking of surgical tools is an XR technological barrier for modeling and intra-operative training. Importantly, the most reported application of XR is VR for simulation-using platforms like the Mentice VIST and Angio Mentor. 10 studies examine VR for simulation, showing enhanced procedural performance and reduced fluoroscopy use after short training, although long-term outcomes have not been reported. Early-stage trainees benefited the most. Simulation improved collaboration between neurosurgeons and the rest of the surgical team, a promising role in interprofessional teamwork. Given the strength of VR for simulation, MR for simulation is an important gap in the literature for future studies. In conclusion, XR holds promise for transforming neurosurgical education and practice for simulation, but technological research is needed in modeling and intra-procedural training.