Abstract: Objective To explore the effectiveness of medical artificial intelligence in long-term health management of type 2 diabetes mellitus patients. Methods A total of 255 T2DM patients who were hospitalized in the endocrinology department of our hospital from January 2020 to December 2021 were randomly divided into three groups: routine group (n=85), control group (n=85) and experimental group (n=85). The routine group only received routine discharge guidance and outpatient interventions, while the control group managed their daily lives and monitored blood glucose through medical artificial intelligence software on the basis of discharge guidance. The experimental group received regular follow-up guidance or outpatient review on the basis of the control group and received comprehensive treatment in five aspects: diabetes education, diet therapy, exercise, blood glucose monitoring, and drug adjustment. Blood glucose control-related indicators, blood lipids, body mass index, self-management capacity and satisfaction were compared among the three groups after 6 months of treatment. Results After 6 months of treatment, the blood glucose (FBG, PBG), glycated hemoglobin (HbA1c), blood lipids, and body mass index of the experimental group were significantly improved compared to those of the control group and the routine group (P<0.05), and the control group was superior to the routine group (P<0.05). Conclusion The Application of medical artificial intelligence in long-term health management of T2DM patients can improve metabolic disorders and weight, increase satisfaction with healthy living and treatment, and improve blood glucose control rate, providing an effective method for chronic disease management.

Key words: type 2 diabetes, artificial intelligence in medicine, blood glucose, health management