Date of Award

Spring 1977

Document Type

Thesis

Degree Name

Master of Science in Electrical Engineering - (M.S.)

Department

Electrical Engineering

First Advisor

Stanley S. Reisman

Second Advisor

Robert De Lucia

Third Advisor

Joseph J. Strano

Abstract

A computer model for the Dynamics of the Glucose-Insulin-Glucagon System has been developed for a 17.5 kg canine using the Continuous Systems Modeling Program (CSMP) for the 360 Computer System. The major body components controlling the glucose dynamics (liver, pancreas, body muscle, blood flow, and body fluid compartments) have been modeled in terms of either their production, absorption, or transport of glucose, and the concentration levels of both the hormones and substrates perfusing the body component. A set of mneumonics has also been developed to label the hundreds of constant and variable terms required to describe a complex system of this magnitude. The dynamic characteristic of the liver's glycogen storage capability has also been modeled in terms of stored glycogen and the blood plasma concentration levels of both glucose and insulin perfusing the liver.

Once the Glucose-Insulin-Glucagon System had been modeled, it was first tested under basal conditions with three different levels of glycogen stored in the liver to check the dynamics of the liver glycogen storage. As expected, when the stored glycogen was below the equilibrium level, blood glucose was converted to liver glycogen, and when the stored level was greater, glycogen was converted back to glucose and returned to the blood.

The Glucose-Insulin-Glucagon System model was then tested with an almost instantaneous glucose load of 8.75 grams of glucose, elevating the glucose concentration level to approximately 3.5 mg of glucose per ml of blood plasma. This high glucose concentration level returned exponentially over the next 120 minutes to the basal concentration level of 100 mg/100 ml, agreeing generally with in vivo test data.

The Glucose-Insulin-Glucagon System model was then tested by injecting insulin into the model at different rates over an extended period of time and observing the rate at which the glucose concentration fell, its final level, and the rate at which the glucose concentration level returned to the basal concentration level once the insulin load had been removed. Here again, there was generally good agreement with in vivo test data, not only for the glucose concentration dynamics but also for the rate at which glucose was produced by the liver during the period when insulin was being injected into the model.

Share

COinS