Artificial Intelligence in Vascular Surgery: Hype, Evidence, and Future Prospects

Abstract

Ahmed Hassan, Aya Mohammed, Mohammed Al Abadie and Sami Al Abadie

Artificial intelligence (AI) is rapidly advancing across surgical disciplines, and vascular surgery is now entering a period of accelerated integration. Recent developments in machine learning (ML), deep learning (DL), and computer vision have improved vascular imaging interpretation, endovascular planning, postoperative surveillance, and wound care. Large registry-based studies published between 2023 and 2025 demonstrate that ML models outperform traditional risk scores in predicting outcomes after endovascular aneurysm repair (EVAR), open and endovascular peripheral artery disease (PAD) interventions, and aortic surgery. DL-based imaging models have achieved high accuracy for endoleak detection, aneurysm segmentation, plaque characterization, and classification of diabetic foot ulcers (DFU), while mobile-enabled AI platforms show promise for remote surveillance and early identification of complications.

Despite this progress, translation into routine vascular practice remains limited. Key barriers include limited external validation, data heterogeneity, model bias, integration challenges, the need for explainability, and evolving regulatory requirements. This review synthesizes the most recent evidence (2023–2025), evaluates clinical readiness across major vascular domains, and outlines ethical, operational, and regulatory considerations. Widespread adoption will require robust multicentre evaluation, multimodal data integration, interoperable infrastructure, and coordinated surgeon–engineer collaboration. AI promises significant enhancements to diagnostic accuracy, risk prediction, and postoperative management, but its safe implementation demands rigorous governance and ongoing validation.

PDF