Abstract:
Rationale: Sheep are increasingly used as a model for human ventilation, however there are substantive anatomical differences between human and ovine lungs that may affect the gravitational distribution of tissue at rest and during ventilation. Understanding ventilation and gas transport in the ovine lung is important for interpreting measurements acquired via in vivo gas contrast imaging, such as xenon enhanced computed tomography (Xe-CT). In this study a computational model that integrates finite elastic deformation of the soft tissue with distribution of inspired air was applied to the ovine lung to determine whether a model that is consistent with human ventilation and gas distribution in also suitable for the simulation of ventilation in experimental animals. Methods: Xe-CT imaging was acquired in three sheep at the University of Iowa Division of Physiologic Imaging. Image based finite element meshes were constructed for each lung and the airway tree at a positive end expiratory pressure (PEEP) of 25 cmH 20 in prone and supine postures. Ventilation and gas transport were simulated for each animal using a computational model that includes finite deformation elasticity to define the pressure volume relationship of the lung, and time-dependent advective and diffusive transport of inhaled Xe gas. Gravitational deformation was simulated prone and supine, and ventilation was simulated for the prone lung. The specific volume change predicted from the advective flow simulation was compared to specific ventilation calculated using the time constant method that is used in Xe-CT analysis. Results: The finite deformation model accurately predicts the regional tissue density in prone but underestimates the supine gradient. Ventilation simulated in the prone posture for each of the three animals indicates that the ventilation distribution predicted by an advective flow model typically differs from that predicted from the time constant method of analysis. The R 2 correlations between the simulations and time constant calculations were 0.7824, 0.3423 and 0.42881 for each of the three animals. Conclusions: The human-consistent model is inadequate for predicting ventilation in supine sheep, because the model does not include movement of the heart or fluid shifts that were evident in the imaging. However the model is sufficiently predictive for the prone ovine lung. That advective ventilation differs from inert gas transport is consistent with experimental findings that show a correlation of R 2 of 0.66 between specific volume and ventilation calculated from tracer gas mixing (1). Fuld et al (2008) J. Appl. Physiol 104:1174-1184 This abstract is funded by: NIH Grant ROI-HL-064368-06A1