Quantifying tissue heterogeneity using quadtree decomposition

Abstract

Volumetric computed tomography (CT) imaging provides a three-dimensional map of image intensities from which lung soft tissue density distribution can be estimated. The information gained from analyzing these images can prove valuable in diagnosis of conditions where lung tissue is damaged or has degenerated, and it is also necessary for modeling lung tissue mechanics. This paper presents a new technique for quantifying heterogeneity based on individual CT images, and investigates the heterogeneity of lung tissue in a group of healthy young subjects. It is intended that development of this technique leads to a standard model of classifying heterogeneity in lung tissue, while taking into account variables such as different imaging platforms and resolutions, and the position of the patient during imaging.