A fuzzy linguistic model for ambulatory systolic blood pressure variation prediction

Establishing relationships between blood pressure variations and body acceleration and heart rate measurements is a key issue to determine the drug quantity to inject to cure patients suffering from hypertension, i.e. high blood pressure. For this purpose, a fuzzy linguistic model is developed from experimental data with a method which leads to a small number of linguistic terms and rules, easily understandable by doctors. The software instrument developed has predictive performances quite similar to those obtained with statistical models. In addition, it provides a unified formalism that will probably simplify the introduction of other features in the drug quantity decision system