Abstract: Hemodialysis is the mainstay of renal replacement therapy for end-stage renal disease, and arteriovenous fistula (AVF) is the preferable method for vascular access recommended by major guidelines. However, repeated AVF failures affect the quality of life of the patients, and increase economic and social burdens. Therefore, continuous assessment of AVF function and early intervention to abnormal AVF is essential. Currently, artificial intelligence has become a hot issue due to the advantages of accurate and quantified results, homogenized and remote diagnosis and treatment, as compared to the physical examination of AVF. In this article, research progresses in AVF acoustic feature and its extraction method, selection of machine learning method, and the development of AVF monitoring system by artificial intelligence are reviewed in order to explore the research pathways and the direction of clinical research.

Key words: Arteriovenous fistula, Artificial intelligence, Machine learning, Audio