Classification of Tree and Network Topology Structures in Medical Images

Several structures of human body follow the topology of a tree or a network. Characteristic examples are the blood vessel network, the bronchial tree, the neuron system, and the breast ductal tree. The analysis of these structures is based on the identification and the quantification of parameters that model and characterize the studying topologies. The objective of this study is the morphological analysis of arterial networks depicted in digital subtraction angiographic images from the well-known hindlimb ischemia model in New Zealand White Rabbits, in order to detect discriminative characteristics of arterial structures in normal limbs and in chronic ischemic limbs after collateral arterial network formation through angiogenesis 40 days after induction of ischemia. Our methodology for characterizing the arterial topologies is based on the analysis of spatial distribution of branching points and the quantification of tortuosity. The experimental evaluation of the approach demonstrates that topological analysis can potentially aid in the discovery of new correlations between morphology and function of the studied structures having as basic application the discrimination of healthy and pathological situations.