Research Of Pulmonary Airway3D Tree Analysis Based On CT Images

With the changes of medical diagnosis and treatment form, medical pattern has changed greatly. One of main changes is the appearance of Computer-aided detection. Respiratory illness is one type of important non-infectious diseases. We use CAD systems in lung disease detection and diagnosis greatly. The research of pulmonary airway3D tree analysis based on CT images is one of important branches in CAD systems.The goal of this research is to realize pulmonary airway3D tree analysis algorithm to exact skeleton, generate a formal tree structure and exact2D cross-section. All the CT images data used in this research come from ShengJing hospital of China medical university. A1thinning algorithm is realized by checking border points of this binary airway tree that satisfy four certain requirements. Then small segments on skeleton are deleted by pruning algorithm. Then tree point, branch points, end points, and line points could be recognized. So, by binary tree topological structure, formal tree structure can be built and the branches of the airway tree also can be labeled in generation order. At last, after smoothing the branch line tangent vector would be calculated, and2D cross-section is achieved by coordinate transforming.Eighteen airway skeletons are exacted successfully using the analysis algorithm proposed in this research. The different threshold in pruning algorithm is satisfied by doctors. A formal pulmonary airway tree structure is achieved and branches are generated. At last, all levels of the bronchial2D cross-section is realized. Pulmonary airway3D tree analysis algorithm in this research can be used to quantitatively analysis and is helpful to pulmonary diseases detection.