[Current applications and challenges of artificial intelligence in diabetes management]

In recent years, the rapid development of artificial intelligence (AI) has brought innovative opportunities to diabetes management, with significant application potential in various aspects such as prevention, screening, diagnosis, and treatment of diabetes. Through big data analysis, AI can achieve precise diabetes sub-typing and further promote personalized treatment. In terms of complication screening, AI can efficiently identify diabetic retinopathy, aiding in early diagnosis and treatment of complications. Combined with continuous glucose monitoring technology, AI can remotely assist in diabetes management through various novel glucose monitoring metrics and algorithms. The AI-based closed-loop insulin pump can automatically adjust insulin infusion, increasing glucose time in range and reducing the risk of hypoglycemia. However, this technological wave is also accompanied by multiple challenges: data privacy disputes, insufficient algorithm interpretability, and delays in clinical validation, which hinder its widespread clinical applications. While AI demonstrates high efficiency in supporting clinical decision-making, complex case management and psychological care should still be dominated by physicians. In the future, the "physician+AI" collaborative pattern will combine humanistic care with technology to optimize diabetes management together.