A numerical study of periciliary liquid depth in MDCT-based human airway models

Periciliary liquid (PCL) is a critical component of the respiratory system for maintaining mucus clearance. As PCL homeostasis is affected by evaporation and mechanical forces, which are in turn affected by various breathing conditions, lung morphology and ventilation distribution, the complex process of PCL depth regulation in vivo is not fully understood. We propose an integrative approach to couple a thermo-fluid computational fluid dynamics (CFD) model with an epithelial cell model to study the dynamics of PCL depth using subject-specific human airway models based on multi-detector row computed-tomography (MDCT) volumetric lung images.

The thermo-fluid CFD model solves three-dimensional (3D) incompressible Navier-Stokes and transport equations for temperature and water vapor concentration with a realistic energy flux based boundary condition imposed at airway wall. A corresponding one-dimensional (1D) thermo-fluid CFD model is also developed to provide necessary information to the 3D model. Both 1D and 3D models are validated with experimental measurements, and the temperature and humidity distributions in the airways are investigated. Correlations for the dimensionless parameters of Nusselt number and Sherwood number are proposed for characterizing heat and mass transfer in the airways. As one of the key applications of the thermo-fluid CFD model, the water loss rates in the both 1D and 3D airway models are studied. It is found that the secondary flows formed at the bifurcations elevate the regional heat and mass transfer during inspiration and hence the water loss rate, which can only be observed in the 3D models. Among the three human airway models studied in both 1D and 3D, little inter-subject variability is observed for the distributions of temperature and humidity. However, the inter-subject variability could be dramatic for the distribution of water loss rate, as it is greatly affected by airway diameter and regional ventilation.

A method is proposed to construct an ion-channel conductance model for both normal and cystic fibrosis (CF) epithelial cells, which couples an existing fluid secretion model with an existing nucleotide and nucleoside metabolism model (collectively named epithelial cell model). The epithelial cell models for both normal and CF are capable of predicting PCL depth based on mechanical stresses and evaporation, and are validated with a wide range of experimental data.

With these two models separately validated and tested, the integrated model of the thermo-fluid CFD model and epithelial cell model is applied to MDCT-based human airway models of three CF subjects and three normal subjects to study and compare PCL depth regulation under regular breathing conditions. It is found that evaporative water loss is the dominant factor in PCL homeostasis. Between three types of mechanical forces, cyclic shear stress is the primary factor that triggers ATP release and increases PCL depth. In addition, it is found that that greater diameters of the airways in the 4th-7th generations in CF subjects decrease evaporative water loss, resulting in similar PCL depth as normal subjects. Under regular breathing conditions, the average PCL depths of normal and CF is around 6 to 7 µm, with mechanical forces play a greater role in regulating CF PCL depth. Comparing to 7.68 µm normal base level (considered as optimum PCL depth), this average PCL depth is about 8 to 21% lower. This might suggest that mechanical forces alone cannot entirely balance evaporative water loss, and other mechanisms might be involved.