Modeling the influence of the pulmonary pressure-volume curve on gas exchange

Current models of lung mechanics and gas exchange act independently to simulate variations in pressure-volume (PV) and ventilation-perfusion (V/Q) properties in the lungs respectively. However, changes in ventilator pressures can cause alveoli recruitment, collapse or over-distension causing V/Q changes in the lungs that are unaccounted for in these models. A compartmental model of the lungs is presented that is based on a physiological interpretation of lung function and simulates each alveolus individually. By combining this model with currently available lung mechanics and gas exchange models, the effect of changing ventilator settings on gas exchange could be simulated. The model is shown to simulate experimentally measured static PV data from an ARDS patient with an accuracy equivalent to that achieved by the sigmoid function. It could enable quantification of variations in V/Q in the lungs and also gives estimates of other physiological lung properties such as lung density and alveoli compliance. The alveoli model offers a physiologically relevant method of simulating the PV relationship in the lungs and its influence of gas exchange.

I. Introduction

Selecting suitable ventilator settings for intensive care patients with respiratory disorders is important for reducing recovery time and minimizing ventilator-associated lung injury (VALI) [1] [2] [3] [4] [5]. This process is difficult, requiring clinicians to quantify the trade-offs between achieving adequate oxygenation of the blood without stressing the lungs and damaging the alveoli. Ventilator settings, such as positive end-expiratory pressure (PEEP) or tidal volume (V_tidal), affect both the exchange of gas between the alveoli and blood, and the lung mechanics, i.e. the relationship between pressures, volumes and flows. Understanding the effect of ventilator settings on both gas exchange and lung mechanics is therefore necessary to assist in choosing optimal ventilator settings.