目的·建立小鼠持续葡萄糖监测（continuous glucose monitoring，CGM）技术，并对其血糖时间序列进行多尺度熵（multiscale entropy，MSE）分析。方法·选取饮食诱导肥胖型（diet-induced obesity，DIO）小鼠（n=3，DIO组）及对照组小鼠（n=3）为研究对象，利用全植入式血糖遥测系统分别收集2组小鼠的血糖及体温数据，取术后第10~14日数据进行分析，并统计系统的记录时间。利用MATLAB R2019b软件对2组小鼠术后第11~17日的血糖时间序列进行MSE分析，计算每个时间尺度上对应的熵值。结果·成功建立了以全植入式血糖遥测系统为基础的小鼠CGM技术。6只小鼠的平均记录时间为（27.3±9.3）d，共获得232 887个血糖数值。DIO组小鼠平均血糖水平为（7.04±0.71）mmol/L，平均体温为（33.34±0.18）℃。与对照组相比，DIO组小鼠血糖时间序列复杂度较低，但差异无统计学意义。结论·成功建立了小鼠CGM技术；MSE分析发现，DIO型小鼠血糖时间序列复杂度降低，可能是其早期糖代谢异常的表现之一。

Objective·To establish continuous glucose monitoring (CGM) in mice and implement multiscale entropy (MSE) analysis of glucose time series. Methods·Diet-induced obesity (DIO) mice (n=3) and control mice (n=3) were selected as the research objects. The blood glucose and body temperature data of the two groups were collected by using the implantable glucose telemetry system. The data of 10?14 days after operation were analyzed, and the recording time of the system was counted. Using MATLAB R2019b software, MSE analysis was performed on glucose time series of the two groups from 11 to 17 days after operation, and the corresponding entropy value on each time scale was calculated. Results·The CGM technology based on the implantable glucose telemetry system in mice was successfully established. The average recording time of the 6 mice was (27.3±9.3) d, and 232 887 blood glucose values were obtained. The mean blood glucose level of the DIO mice was (7.04±0.71) mmol/L and the mean body temperature was (33.34±0.18) ℃. Compared with that of the control mice, the glucose time series complexity of the DIO mice was lower, and the difference between the two groups was not statistically significant. Conclusion·The CGM technology in mice is successfully established. MSE analysis shows that the complexity of glucose time series in the DIO mice decreases, which may be one of the manifestations of abnormal glucose metabolism in the early stage.