Abstract
As a tool for simulating the variation of blood pressure of supine patients caused by acceleration and deceleration of the ambulance, we made a mathematical model for estimating the blood pressure variation by using the system identification technique. Prior to the modeling, to make a statistical analysis of the relationship between the foot-to-head acceleration and the blood pressure variation, we recorded the blood pressure for 15 healthy male subjects at the age of 21-24 years by exposing them to two different patterns of acceleration in foot-to-head direction with the tilt bed. It was found that the mean blood pressure, which was measured at the left finger tip put on the left side of the head, varied proportionally to the acceleration on average. Based on these experimental data, we constructed a probability model of the blood pressure variation. The parameter values of the model were determined optimally by the genetic algorithm so that two mean-squared-error values, which were calculated for each pattern of acceleration, were minimized in the sense of Pareto optimality. The validity of the constructed model was confirmed by the car experiment.
