Anatomical model matching with fuzzy implicit surfaces for segmentation ofthoracic volume scans

Many segmentation methods for thoracic volume data require manual input in the form of a seed point, initial contour, volume of interest etc. The aim of the work presented here is to further automate this segmentation initial ization step. In this paper an anatomical modeling and matching method is proposed to coa rsely segment thoracic volume data into anatomically labeled regions. An an atomical model of the thorax is constructed in two steps: 1) individual org ans are modeled with blended fuzzy implicit surfaces and 2) the single orga n models are grouped into a tree structure with a solid modeling technique named constructive solid geometry (CSG), The combination of CSG with fuzzy implicit surfaces allows a hierarchical scene description by means of a bou ndary model, which characterizes the scene volume as a boundary potential f unction. From this boundary potential, an energy function is defined which is minimal when the model is registered to the tissue-air transitions in th oracic magnetic resonance imaging (MRI) data. This allows automatic registr ation in three steps: feature detection, initial positioning and energy min imization, The model matching has been validated in phantom simulations and on 15 clin ical thoracic volume scans from different subjects. In 13 of these sets the matching method accurately partitioned the image volumes into a set of vol umes of interest for the heart, lungs, cardiac ventricles, and thorax outli nes. The method is applicable to segmentation of various types of thoracic MR-images, provided that a large part of the thorax is contained in the ima ge volume.