An interactive approach to liver segmentation in CT based on deformable model integrated with attractor force

The segmentation of the liver in CT (Computed Tomography) volumetric data provide a basis for constructing the three dimensional anatomical model, which is an important prerequisite for a variety of clinical and research applications. In this paper, we introduce an algorithm for semi-automatic liver segmentation in three dimensional CT images based on the deformable model. The deformable model is initialized as an existing simplex mesh model of the liver, which was created from manual segmentation. Deformation of the shape model to adapt to the boundary of a new liver is formulated by minimizing a local cost function associated with each model vertex. In particular, we introduce a new attractor force component to constrain the mesh deformation and refine the segmentation result. The experimental results demonstrated the effectiveness of the proposed method.