A355 Pushing the frontiers of neurointervention: a systematic comparison of robotic-assisted and manual endovascular techniques

<h3>Introduction</h3> Robotic neurointervention is redefining the landscape of endovascular therapy and care. Yet, its clinical utility across cerebrovascular indications remains incompletely defined. <h3>Aim of study</h3> To evaluate procedural efficiency in performance, patient safety, and radiation exposure across robotic-assisted and manual interventional neuroradiology (INR) procedures in a stratified, indication-specific analysis. <h3>Method</h3> A rigorous PRISMA-compliant review (Pubmed, Embase, Medline, Scopus, Web of Science, Cochrane; 2000–2025) identified 400 records, with 13 studies (n=538 procedures, ≥4 patients/study) matching inclusion criteria. Both diagnostic and therapeutic interventions across all ages and pathologies were analysed. Outcomes: technical and clinical success, complications, radiation exposure, procedural time, and conversion to manual technique. <h3>Results</h3> Technical success for robotic procedures reached 93% (range 88–99%), with a minimal complication rate (0.7%) and zero mortality. Manual conversion was required in 7%, primarily due to anatomical or device limitations. All studies reported some manual steps, reflecting current robotic constraints. Subgroup analysis (refer table 1) confirmed robust outcomes across adults, pediatrics, aneurysm, stroke, and both diagnostic and therapeutic procedures. Notably, operator radiation exposure was reduced by up to 90% in robotic cases. Only one comparative study found longer procedural times for robotics, but no inferiority in safety or efficacy. <h3>Conclusion</h3> While robotic-assisted INR is not yet autonomous, it delivers competitive procedural efficacy with significantly reduced operator radiation thereby promising futuristic innovations. Unlike previous reviews, our analysis dissects outcomes by procedural type in adults as well as paediatric populations and highlights underreported system limitations. These insights support targeted adoption and inform future robotic design and randomized evaluation. <h3>Conflict of Interest</h3> No