Assessment of arterial gas pressures and cardiac output using a breathing lung model

PURPOSE: The aim of this investigation was to evaluate a breathing lung model to estimate arterial gas partial pressures and cardiac output at rest and during exercise. METHODS: A mathematical model was used to describe variations in gas fractions, alveolar volume, and gas exchange in the pulmonary capillaries during the breathing cycle. Experimental data were obtained from 17 healthy subjects at rest and during exercise at 40, 50, 65, and 75% VO(2max) on a cycle ergometer. VO(2), VCO(2), and P(ET,CO2) were monitored continuously with a MedGraphics CPX/D gas exchange system. Arterial gases were measured in brachial artery blood samples drawn simultaneously with gas exchange. Cardiac output was measured using the CO(2) rebreathing method corrected by the arterial blood sample data. The model parameters including cardiac output, end-expiratory alveolar volume, and mixed-venous gas contents were estimated by fitting modelVCO(2) to experimental values over 50 breaths. RESULTS: The fit of model parameters gave arterial gas partial pressures not significantly different from measured data. Measured P(a,C02) and P(a,O2) were significantly correlated with model outputs (R(2) = 0.991 for P(a,CO2) and R(2) = 0.999 for P(a, CO2); P < 0.0001). The cardiac output values estimated using the breathing lung model were significantly correlated with the values obtained with the corrected CO rebreathing method (R(2) = 0.71; P < 0.0001). There was, however, a significant 2.3 L x min(-1) difference between these two methods. CONCLUSION: Results obtained with the proposed method were in good agreement with measured arterial gas partial pressures. Despite a certain degree of bias, the promising results for cardiac output demonstrate the reliability of this method that should be now evaluated using a gold standard method.