维持性血液透析(maintenance hemodialysis,MHD)是终末期肾病(end stage renal disease,ESRD)患者最主要的治疗方式之一，但存在并发症多、死亡率高等问题。临床上有强化MHD相关并发症风险预测与透析全面管理等方面的需求。机器学习(machine learning，ML)凭借数据挖掘与模型构建能力，已逐步在MHD领域开展应用探索。本文围绕ML模型在MHD领域的应用展开综述，重点梳理模型在患者监测、并发症预测及综合管理方面的研究与应用现状，旨在为该类模型在MHD患者临床实践中的进一步推广与优化提供参考。

Maintenance hemodialysis (MHD) is the primary treatment for end-stage renal disease (ESRD). MHD patients often experience multiple complications with a higher mortality rate. Clinically, there is a need for enhanced prediction of MHD-related complication risks and comprehensive management of hemodialysis. Machine learning (ML), leveraging its data mining and model-building capabilities, has gradually been explored for application in MHD field. This review focuses on the application of ML models in MHD, highlighting the current research and practical use of these models in patient monitoring, complication prediction, and comprehensive management. It aims to provide a reference for further promoting and optimizing such models in clinical practice of MHD.