Segmentation of human upper airway using a level set based deformable model
Xianghua Xie · 2009
In this paper, we present a preliminary study on segmenting a human upper airway from a 3D CT scan using a level set based deformable surface model. The human upper airway has a very complex geometry and its topology may vary from individual to individual. Accurate 3D geometry reconstruction is essential in understanding airway disease and a prerequisite for patient-specific computational fluid dynamics analysis. The proposed method uses a hypothesized dynamic interaction force between the deformable surface and object boundaries which can greatly improve the deformable model performance in acquiring complex geometries, boundary concavities, and in dealing with weak image edges. The results show that the proposed deformable model can be used to efficiently segment complex and compact structures such as the nasal cavity from a 3D image dataset.