Top Things to Know: Artificial Intelligence in Peripheral Artery Disease
Updated: May 14, 2026
Prepared by Barbara Entl, MD, Science and Medicine Advisor, American Heart Association
- Peripheral artery disease (PAD) is common but often underdiagnosed, as many patients are asymptomatic and disparities in access to care limit detection. These challenges highlight the need for innovative approaches, including artificial intelligence (AI), to enhance diagnosis, risk prediction, and management.
- Rapid advances in AI enable high-throughput analysis of complex vascular data, such as arterial waveforms, cardiovascular imaging, and clinical variables, creating opportunities to transform vascular care through improved diagnostics, risk assessment, and therapeutic planning.
- AI supports disease detection, phenotyping, diagnostic imaging, clinical decision-making, and perioperative planning, enhancing visualization and quantification of blood flow across CT, MRI, and ultrasound.
- In this context, this science advisory provides guidance for clinicians, researchers, and policymakers on the responsible, evidence-based deployment of AI in PAD diagnosis and management.
- Supervised machine learning and deep learning models show promise in predicting adverse PAD outcomes, including MACE (major adverse cardiovascular events) and MALE (major adverse limb events), post-procedural complications, and mortality, but further validation and implementation studies are needed before routine clinical adoption.
- Machine learning can also help identify undiagnosed aortic aneurysms, detect them in arterial imaging, and predict growth or rupture, with AI-driven analysis improving accuracy, especially for small or anatomically complex aneurysms.
- AI systems may lose accuracy when applied to different patient populations, workflows, or technologies than those used in training, and algorithmic bias from non-representative datasets can further limit reliability and generalizability.
- Clinician familiarity and confidence with AI tools are critical for safe and effective integration into diagnostic and therapeutic workflows as these systems become increasingly embedded in vascular care.
- Key barriers to implementation include limited AI training, data quality concerns, legal and regulatory uncertainty, unclear reimbursement pathways, and challenges related to trust, explainability, and integration into existing clinical workflows.
- AI involves multiple stakeholders, including clinicians, developers, and institutions; for now, clinicians are likely to remain ultimately responsible for patient care decisions as legal and professional frameworks for AI in healthcare evolve.
Citation
Harzand A, Ross EG, Weissler EH, Zheng Y, Shah NH, Alabi O, Attia ZI, Beckman JA; on behalf of the American Heart Association Council on Peripheral Vascular Disease; Council on Cardiovascular and Stroke Nursing; Council on Cardiovascular Radiology and Intervention; Council on Clinical Cardiology; and Council on Quality of Care and Outcomes Research. Artificial intelligence in peripheral artery disease: a science advisory from the American Heart Association. Circ Popul Health Outcomes. Published online May 14, 2026. doi: 10.1161/HCQ.0000000000000146