Non-uniform central airways ventilation model based on vascular segmentation

G Ibrahim,A Rona,S.V Hainsworth

doi:10.1016/j.compbiomed.2015.08.006

Abstract

Improvements in the understanding of the physiology of the central airways require an appropriate representation of the non-uniform ventilation at its terminal branches. This paper proposes a new technique for estimating the non-uniform ventilation at the terminal branches by modelling the volume change of their distal peripheral airways, based on vascular segmentation. The vascular tree is used for sectioning the dynamic CT-based 3D volume of the lung at 11 time points over the breathing cycle of a research animal. Based on the mechanical coupling between the vascular tree and the remaining lung tissues, the volume change of each individual lung segment over the breathing cycle was used to estimate the non-uniform ventilation of its associated terminal branch. The 3D lung sectioning technique was validated on an airway cast model of the same animal pruned to represent the truncated dynamic CT based airway geometry. The results showed that the 3D lung sectioning technique was able to estimate the volume of the missing peripheral airways within a tolerance of 2%. In addition, the time-varying non-uniform ventilation distribution predicted by the proposed sectioning technique was validated against CT measurements of lobar ventilation and showed good agreement. This significant modelling advance can be used to estimate subject-specific non-uniform boundary conditions to obtain subject-specific numerical models of the central airway flow.

Highlights

Airflow analysis underpins the understanding of lung physiology and pulmonary target therapy
This paper introduces a lung volume sectioning technique based on vascular segmentation to estimate the non-uniform ventilation within the peripheral airways that are not captured by currently available 4DCT imaging techniques
The vascular tree was used as a reference to section the lung volume generated from successive 4DCT images that sample the breathing cycle into several segments, where each segment encloses the missing peripheral airways associated to each CT captured terminal airway branch

Summary

Introduction

Airflow analysis underpins the understanding of lung physiology and pulmonary target therapy. De Backer et al [2] and Lambert et al [3] adopted a similar iterative approach based on CT measurements of lobar growth rather than of lung growth Such iterative approaches improved the reliability of the numerical predictions of the central airway flow, they are not adequate to accurately capture the non-uniform inner lobar flow distribution. Yin et al [4, 5] used a 3D-1D coupling model associated with a ventilation map computed by registering CT images acquired at different inflation levels to estimate the flow at the terminal branches based on mass conservation Their technique resulted in a good approximation to the CT measurement of lobar ventilation. The apportionment of the lobar ventilation among its terminal branches has to be assumed, which may increase the uncertainty of the predicted ventilation within each lobe

Methods

Results

Discussion

Conclusion