STRATEGIES FOR INSULIN DOSAGE ADJUSTMENT USING MODEL-BASED BLOOD GLUCOSE PREDICTION
Charles Universily · 1990
This paper discusses the planning strategies for modelbased systems which assist in insulin therapy adjustment. Two therapy modes are considered: daily and follow-up therapy adjustment. The properties of a criterion to describe the control of blood glucose are given and a penalty function in the form of M-value is recommended. Definition of an optimum therapy is given using the concept of basal and correction therapy. Four methods for assessment of the basal therapy advice are considered: non-linear regression analysis, A' algorithm of artificial intelligence, simulated annealing algorithm and a linear programming method. A strategy for optimum therapy assessment is discussed and the integration with a knowledge-based system is considered.