Development of a Blood Glucose Control System for Postoperative Diabetics

2003 Fiscal Year Final Research Report Summary

• Project/Area Number: 14550444
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Control engineering
• Research Institution: KYOTO UNIVERSITY
• Principal Investigator: ARAKI Mituhiko Kyoto University, Graduate School of Engineering, Professor, 工学研究科, 教授 (60026226)
• Co-Investigator(Kenkyū-buntansha): FURUTANI Eiko Kyoto University, Graduate School of Engineering, Associate Professor, 工学研究科, 助教授 (40219118) ; ONODERA Hisashi Kyoto University, Graduate School of Medicine, Associate Professor, 医学研究科, 助教授 (50240825)
• Project Period (FY): 2002 – 2003
• Keywords: postoperative diabetes / insulin / blood glucose / modeling / state-predictive control / auto-tuning

Research Abstract

Today, surgical operations on elderlies and diabetics are often performed. After surgical operations, they would sometimes come down with hyperglycemia, which is called "postoperative diabetes." Since this might result in infection, anastomic failure, or other complications, insulin is infused by medical staffs. The infusion rate of insulin has to be accurately adjusted, because patients would fall into hypoglycemia if an excessive amount of insulin is infused, and because blood glucose level would not become low enough if a sufficient amount of insulin is not infused. However, it is not easy to manipulate the insulin infusion rate to maintain the blood glucose within the desired range because the response of blood glucose to insulin is widely varying during a few days after surgery.
In this research, we have studied 1)the model of the blood glucose regulatory system of postoperative diabetics, and 2)a blood glucose control method for postoperative diabetics. The obtained results are as follows.
●From the results of animal experiments, Karlsburg model is the most appropriate model of blood glucose regulatory system in the time-invariant case among the representative models proposed in the past.
●The blood glucose response of postoperative diabetics can be modeled as a second order state equation with a dead-time.
●The change of the blood glucose response with time to insulin can be represented as the change of the gain of the blood glucose regulatory system.
●The state-predictive controller with an auto-tuning function developed for blood glucose control of diabetics without a knowledge of individual parameters is effective as the controller for postoperative diabetics.