论文部分内容阅读
采用多元时序预测模型对目前家用血糖仪中保存的糖尿病患者所测量的血糖水平数据进行分析,研究了不同因素如血糖水平、胰岛素剂量、低血糖的发生对患者血糖水平的影响,提出了一种基于不同时间点的多元时序血糖水平预测模型。实验结果表明该模型对患者的血糖水平预测效果较好,通过一定转换即可获得患者在不同时间点下适合注射的胰岛素剂量。该算法可用于辅助治疗胰岛素依赖型糖尿病患者,尽量避免因不合理用药导致其他疾病的发生。
The multivariate time series prediction model was used to analyze the data of blood glucose level measured by the diabetes patients saved in the current home glucose meter. The effects of different factors such as blood glucose level, insulin dose and hypoglycemia on the blood glucose level of the patients were analyzed. Multivariate sequential blood glucose level prediction model based on different time points. The experimental results show that the model of the patient’s blood glucose level prediction is better, through a certain conversion can be obtained at different times the patient suitable for injection of insulin dose. The algorithm can be used to assist in the treatment of patients with insulin-dependent diabetes mellitus, try to avoid the irrational use of drugs lead to other diseases.