Intensive Care Unit




In recent years in view of the intensive care unit patients blood sugar to strengthen control technology is becoming more and more attention, and began to widely used in clinical. In clinical Intensive Care (Intensive Care Unit, or ICU, the existence of patients with stress hyperglycemia is common, it is in severe cases the body in order to be able to through the crisis period and transfer the stress reaction of the body organs of a short-term high blood glucose response (blood glucose level > 180 mg/dL). Stress hyperglycemia is one of the important hazards of critically ill patients, it can reduce the body's immune function, increase the body's energy metabolism disorder. In addition, a variety of microorganisms, easy to grow in the high blood glucose under the environment of high blood sugar there has been a marked increase in the susceptibility to infection in postoperative patients. Will, therefore, to critically ill patients with stress hyperglycemia and plummeted, and keep in a safe range has become an effective treatment for severe patients, and reduce the mortality of patients with an important measure.
    In order to effectively control blood sugar in patients with continuous monitoring of blood glucose levels and continuous intravenous insulin is two important means. Most of the patients in the rescue process, should know how to use a certain amount of cortical hormone, however these hormones may increase blood glucose; Therefore, it is necessary to continuously monitor the patient blood sugar. Blood sugar levels not only reflects the seriousness of the stress state of patients, more important is the blood sugar level have a direct impact on patients postoperative rehabilitation. Since patients, on the other hand, a certain degree of low blood pressure, edema, low body temperature can cause skin and muscle of blood vessels constrict, absorb bad cause insulin, the subcutaneous insulin injections efficiency decline, therefore, by continuous insulin intravenous drip of insulin pump is the best choice.
    Insulin infusion can affect the patient's blood sugar levels and blood glucose levels, in turn, can also be used as the basis for adjustment of insulin infusion speed. From the point of view of the automation, which is a typical feedback process, therefore, automatic control theory and technology can play an irreplaceable role in this field. This group based on subspace identification is proposed and two kinds of model predictive control method and particle swarm optimization to automatically design insulin infusion speed, realize the purpose of strict control blood sugar levels, keeping the blood glucose of intensive care patients in a safe area and reduces the mortality of the patients and improve their prognosis.
Closed-loop control algorithm for a system design usually consists of two steps, first of all, through the identification or other method based on object model, then based on the identification model of the control law is designed. Due to the diversity of population, physical dynamic reaction process between different individuals often are quite different, for example, different individual's weight can be ranged from 40 kg to 120 kg; Different people are great differences of insulin sensitivity, etc. Therefore, in order to be able to design a good insulin infusion strategy, we need to create a different individual models for different patients. In reality, however, the vast majority of cases, it is not possible, because many patients happened because of an emergency are fed into the intensive care unit (ICU), before that we have almost no on the patients know that nature can't patients individualized model is set up. In addition, the patient was admitted to intensive care room once need to start to control blood sugar, the urgency of the time does not allow us to patients individualized model is obtained by open loop identification. Lack of individualized model for ICU patients blood sugar closed-loop control brought a big challenge.
    "Insulin and blood sugar concentration" is a typical nonlinear dynamic process, and contains a lot of delay. Even if, for example, is using a short-acting insulin, insulin infusion to work usually have 10-15 minutes delay between; Measurement of blood glucose concentration also there is a big delay, compared with venous blood glucose values fingertip blood glucose values tend to have certain hysteresis. At the same time, in this process there are a lot of external disturbance, such as stress, exercise and diet. A large number of clinical practice shows that pressure, exercise and diet is the main factor that affect blood sugar levels several; , however, they are difficult to be real-time measured, so can only be as external disturbance in the control process to deal with.
    In summary, intensive care patient blood sugar closed-loop control can be refined for a nonlinear system with time delay and external disturbance, no model control problem, the problem is one of the important challenges the control theory field, there is no general solution. Physi CAL method is proposed to use subspace method and particle swarm model identification of the patient, after pass the model to the model predictive control to control the blood sugar, after a series of simulation test, to verify the effectiveness of the two methods, our proposed method of blood glucose control in the intensive care patient's glycemic control provides a new effective way.



Copyright @ 2013 AutomMed. All rights reserved. Powered by Doccms  京ICP备13030616号