Experimental Parameterization of a Model of Hypoxia Dynamics in Yorkshire Swine

This paper estimates the parameters of a model of hypoxia in large animals (specifically, Yorkshire swine) from experimental data. The paper adopts a simple, control-oriented model of oxygen transport and consumption dynamics. The model uses three state variables to represent oxygen concentrations and/or partial pressures in the animal's alveolar, arterial, and venous compartments. Hypoxia is induced in a laboratory animal by placing the animal on mechanical ventilation and reducing the fraction of inspired oxygen. The resulting dataset is used for the optimal estimation of key parameters governing the animal's oxygen transport, oxygen consumption, and hemoglobin dissociation. In this nonlinear system identification exercise, the optimization objective is a weighted sum of the mean square errors in predicting (i) blood oxygen saturation and (ii) end-tidal oxygen fraction. The resulting model fits the experimental hypoxia data reasonably well, and is intended as a baseline for ongoing research on the treatment of respiratory failure.