Artificial intelligence-enhanced vascular access in hemodialysis – challenges and innovations

Authors

Zayan, Tariq A.; Khafagy, Heba A.

Citation

Arab J Nephrol Transplant. 2026;8(1):11–23.

Abstract

Background: Vascular access represents the lifeline for patients undergoing hemodialysis, yet its management faces dual challenges: a global shortage of specialized surgical workforce and high rates of access dysfunction. This review examines the intersection of artificial intelligence (AI) applications in vascular access surveillance and innovative workforce solutions addressing global surgical shortages. Methods: A comprehensive narrative review was conducted synthesizing evidence from randomized controlled trials, observational studies, and international guidelines. Literature was identified from PubMed, Scopus, and Web of Science, with emphasis on publications from 2020 to 2025 addressing AI-based vascular access surveillance algorithms and task-shifting workforce models in hemodialysis. Results: Machine learning algorithms demonstrated promising results in predicting arteriovenous fistula and graft stenosis through analysis of hemodynamic parameters, ultrasound imaging, and electronic health record integration. Convolutional Neural Networks and deep learning models achieve sensitivity and specificity exceeding 90% in detecting early access dysfunction, potentially reducing thrombosis events and catheter dependence. Concurrently, the global nephrology community confronts critical shortages of vascular surgeons, particularly in low-income and middle-income countries, where permcath dependence rates exceed 70% in some regions. Task-shifting models involving nephrologist-performed catheter insertions and maintenance procedures demonstrated safety profiles comparable to traditional surgical approaches, with technical success rates of 95-98%. Conclusions: An integrated framework combining AI-enhanced surveillance with workforce development strategies offers a promising pathway to optimize vascular access outcomes. Implementation challenges in resource-limited settings require scalable technologies, adapted training models, and supportive policy. As the global burden of end-stage kidney disease continues to rise, this dual approach represents a transformative strategy to reduce morbidity associated with access dysfunction.

Full text (excerpt)

Artificial intelligence-enhanced vascular access in hemodialysis - challenges and innovations Tariq A. Zayan, Heba A. Khalagy Division of Nephrology, Sur Hospital, Sur, Oman. Correspondence to Dr. Tariq A. Zayan, FRCP, Division of Nephrology, Sur Hospital, Sur, South Ash Sharqiyah 411, Sultanate of Oman. Postal code: 411. Tel: +968-95956956/77033; E-mail: [email protected] Received 11 October 2023 Revised 01 November 2023 Accepted 04 November 2023 Published 01 April 2026 Arab Joul of Nephrology and Transplantation 2026; 8:11-23 Background Vascular access represents the lifeline for patients undergoing hemodialysis, yet its management faces global challenges. A global shortage of a specialized surgical workforce and high rates of access dysfunction. This review examines the intersection of artificial intelligence (AI) applications in vascular access surveillance and innovative workforce solutions addressing global surgical shortages. Methods A comprehensive narrative review was conducted synthesizing evidence from randomized controlled trials, observational studies, and international guidelines. Literature was identified from PubMed, Scopus, and Web of Science, with emphasis on p

Read More in AJNT

The Arab Journal of Nephrology and Transplantation is the peer-reviewed open-access journal of ASNRT, published since 2008.

Issue table of contents — Vol 8, Issue 1 · AJNT archive · Download full-text PDF (open access)