An anatomically based hybrid computational model of the human lung and its application to low frequency oscillatory mechanics.

Lung input impedance measured via forced oscillation over low frequency range has been confirmed as sensitive to the degree and the heterogeneity of lung disease. In this study we advanced an image-based, multi-scale computational model for the human lung, which includes upper and central airways, small airways and alveoli tissue unit. A three-dimensional (3-D) realistic model of the upper airway (reconstructed from MRI images) was combined with an anatomically based 3-D model of the central airways (based on MDCT images) to form a 3-D model of the large airways (from mouth to generation 6, incomplete for generations 4-6). The small airway trees distal to the central branches were based on a hypothetical airway tree for a normal healthy lung. A constant phase viscoelastic model was assumed for the alveolar tissue unit. Unsteady airflows in the large airways were simulated based on computational fluid dynamics (CFD). An experimentally measured broadband forcing flow was applied at the mouth. The impedance of the small airways was computed based on a one-dimensional transmission line model. The computed overall dynamic lung resistance and elastance compared very well with experimental values. Results showed that unsteady 3-D simulation and realistic geometry of the upper and large airways up to generations 4-6 can provide a reasonably accurate estimation of lung input impedance. The impedance of the upper airway constitutes a significant part of the total lung input impedance. The resistance of the upper airway accounts for 45-70% of the total lung resistance at frequencies between 0 and 1 Hz, and 70-81% at frequencies between 1 and 8 Hz.