Pipeline for 3D Reconstruction of Lung Surfaces Using Intrinsic Features Under Pressure-Controlled Ventilation

Abstract

The measurement of whole lung mechanics forms the basis of diagnostic measurements for many respiratory diseases. Despite this, there are currently no quantitative methods to link alterations in pulmonary microstructures to measurements of whole lung function. The normal decline in the lung’s microstructure that occurs with age is virtually indistinguishable from early disease on imaging or standard lung function measurements, leading to frequent misdiagnosis in the elderly. Accurate characterisation of lung mechanics across spatial scales has the potential to assist distinguishing age from pathology, which would benefit patients across a range of medical conditions and procedures. While computational modelling promises to be a useful tool for improving our understanding of lung mechanics, there is currently no unified structure-function computational model that explains how age-dependent structural changes translate to decline in whole lung function. This paper presents novel instrumentation and imaging techniques for measurements of intact ex vivo lung tissue mechanics. We seek to address problems of weak parameterisation that existing models suffer from, due to lack of reliable measurements. To begin addressing this issue, we have developed a full-field stereoscopic imaging system for tracking surface deformation of the rat lung during pressure-controlled ventilation. This study presents a pipeline for the reconstruction and tracking of the intact left lobe of a rat lung during inflation, ex vivo. Model-based 3D reconstruction of the lungs enabled the 3D shape of a surface patch of the imaged lung to be determined. The 3D reconstruction and tracking of the fresh lung surface patch in this study was completed with three cameras across 21 pressure steps, encompassing a total pressure change from 2069 Pa to 2386 Pa. This approach shows that reconstructing intact ex vivo fresh lungs, with no additional surface markers, is feasible.