Abstract:
Rationale: The conductive index that is derived from the multiple-breath inert gas washout test (Scond) is thought to be indicative of regional ventilation inhomogeneity. Scond is primarily of interest in studying airway disease, however the index results from a combination of heterogeneity induced by the airways and the regional compliance of the parenchyma. Xenon-CT imaging provides a measure of regional ventilation. Scond however is critically dependent upon sequential ventilation which is not measured in the multi-breath Xenon-CT method. Methods:We simulate single and multiple inert gas washout tests in models of healthy human lungs to determine whether the distribution of tissue density measured by CT can be used to simulate realistic nitrogen washouts in upright and supine postures. Time-dependent ventilation distributions are simulated in an anatomically-based model of the upright human lung derived from MDCT imaging. Airway measurements, and density distributions of the compliant alveolar tissue have been validated against CT imaging. Equations for Poiseulle flow, additional energy dissipation due to the airway bifurcations, conservation of mass, and the equation of motion are solved to produce a ventilation distribution. This is subsequently used in simulation of inert gas mixing. A distribution of end expiration regional time constants is concurrently produced. Results: Model behavior was verified by comparison with published experimental results for Scond and gas bolus dispersion. The phase III slopes of 150 ml boluses of inert gases inhaled at 0, 20, 40, 60 and 80 % of vital capacity are consistent with experimental trends, producing a positive slope for boluses inhaled early in the breath, and a negative slope for boluses inhaled later in the breath. This confirms that sequential ventilation occurs in the model. Tissue density is distributed in an isogravitational plane by soft tissue mechanics. In the absence of compliance heterogeneity airway heterogeneity does not produce significant ventilation heterogeneity in the model. Uniform airway constriction results in a decreased Scond indicating that bronchomotor tone could influence the regional distribution of ventilation in healthy subjects. Localized airway constriction leads to increased isogravitational heterogeneity and an increased Scond. Conclusions: Sequential ventilation of the non-linear lung tissue results in a positive Scond of realistic magnitude. The generation of Scond is dependent upon sequential ventilation due to compliance heterogeneity and is modified by the airway geometry. The model provides a first structure for linking the mechanisms underpinning washout tests at the mouth with regional Xenon-CT imaging. This abstract is funded by: NIH Grant RO1-HL-064369-06A1