Abstract: The prevalence of end stage renal disease (ESRD) has been steadily increasing, and hemodialysis is a predominant treatment modality for these patients. Recently, a series of artificial intelligence (AI) models including machine learning and deep learning techniques have been successfully developed into the area of hemodialysis research. Compared with the conventional linear models, the AI models can accurately predict complications, evaluate vascular access, manage volume control and estimate prognosis based on the information from hemodialysis patients. AI also has powerful potentials in clinical practice to improve and individualize hemodialysis protocols and to prevent and treat complications. This review summarizes recent the research advances in AI models for monitoring hemodialysis and predicting complications, aiming to support doctors and nurses to improve the quality of life and to prolong the lives of hemodialysis patients.

Key words: Artificial intelligence, Hemodialysis, Machine learning, Deep learning