Development of the short-term blood glucose prediction algorithm for using in closed-loop insulin therapy device

Abstract

Algorithm for short-term blood glucose values prediction was developed. Based on mathematical model of blood glucose dynamics it classifies measured BG values, calculates bolus insulin values, generates alarm signals for the patient notifying. It also detects critical blood glucose values and errors and faults of glucometer and insulin pump contained in the artificial pancreas device. Testing results make possible doing further more complex steps to implement the closed-loop insulin therapy device for clinical use.