AI-driven decision making for intravascular device selection in aortic disease. Current insights and prospects

Abdominal and thoracic aortic repairs increasingly rely on endovascular solutions, but device selection in anatomically complex cases remains prone to error due to measurement variability, tortuosity, short/angulated necks, and heterogeneous post-EVAR evolution. This article focuses on artificial intelligence (AI) tools that support intravascular device selection and planning, particularly in abdominal and thoracic aortic aneurysms, and type B dissection scenarios where endovascular repair (EVAR/TEVAR) is applicable. We synthesize evidence on (i) automated centerline extraction and 3D measurements that standardize sizing; (ii) risk models that predict inadequate sealing or endoleakage to guide oversizing and landing zone strategy; and (iii) procedural environment "augmented intelligence" maps and extended reality modules that operationalize device selections in the laboratory. We summarize commercial and research-level systems, clinical readiness, and regulatory status, and outline validation, explainability, and bias considerations. While current evidence-based workflows achieve excellent results, targeted AI components reduce variability and can support consistent device decisions across complex anatomies. Prospective, multicenter validation is needed before routine implementation; for now, AI should be viewed as a complement, rather than a replacement, to established EVAR/TEVAR planning and oversight.