Artificial Intelligence‐Enhanced, Closed‐Loop Wearable Systems Toward Next‐Generation Diabetes Management

Recent advancements in wearable healthcare have led to commercially accessible continuous glucose monitoring systems (CGMs) for diabetes management. However, CGMs only monitor glucose levels and lack therapeutic functions, prompting the development of closed‐loop systems that use monitored glucose levels to guide insulin dosing. While promising, these devices also pose risks, such as insulin overdosing, which can cause hypoglycemia. This review summarizes recent advances in integrating artificial intelligence methods with conventional CGMs. The developments in wearable CGMs and progress in insulin delivery technologies are explored, and existing algorithms for glucose prediction in closed‐loop systems are reviewed. Additionally, emerging trends in optimizing these algorithms to enhance the safety and security of closed‐loop insulin delivery systems are highlighted.