Modeling of the Heart Rate Variation Caused by Acceleration and Deceleration of an Ambulance

In emergency transportation by an ambulance, the inertial acceleration acts on an injured or sick patient in the craniocaudal direction when the ambulance accelerates or decelerates. The acceleration fluctuates the blood pressure of the patient unnaturally. To control the fluctuation, the arterial baroreceptor reflex changes the heart rate as well as the peripheral vascular resistance. The heart rate change might give a bad effect to a patient with heart disease. The purpose of this paper is to construct a computer simulation model of the heart rate variation for the risk assessment of the acceleration of the ambulance. The model is given in the form of a normally distributed stochastic model with the standard deviation of 4.1 mmHg. Its mean is estimated by the feedback neural network, which is trained based on the experimental data obtained with the computer-controlled tilt bed for 15 healthy subjects in their twenties. The trained network has a high estimation accuracy: The RMSE and the correlation coefficient between the arithmetic mean of the heart rate variation of 15 subjects and the network output are 0.56 mmHg and 0.78, respectively, for the validation data set. The validity of the model is examined for the car experiment. The frequency response of the network is also shown.