Abstract
To date, there has been no standard method for predicting blood glucose in clinical use. In this paper, we discuss the development of a new controlling method that can improve the hyperglycemia state in Type II diabetes. Since the fasting blood glucose level is regulated by the balance of insulin secretion and resistance, a chaotic model is created by embedding the theorem of Takens from a stabilized time series of the fasting blood glucose level incorporating the mathematical model of the glucose-insulin metabolism. The fasting blood glucose level of the next morning can be predicted using a semi-fuzzy reconstruction method, and the optimal amount for the insulin injection before going to bed can also be calculated from the prediction. As a result, patients can regulate their lifestyle using this quantitative perspective. Under this regulation, both the chaotic model and the prediction become more stable and accurate over a long period of use.
