Automated segmentation of the liver from 3D CT images using probabilistic atlas and multi-level statistical shape model

摘要: An atlas-based automated liver segmentation method from 3D CT images is described. The utilizes two types of atlases, that is, the probabilistic atlas (PA) and statistical shape model (SSM). Voxel-based with PA firstly performed to obtain a region, then obtained region used as initial for subsequent SSM fitting images. To improve reconstruction accuracy especially largely deformed livers, we utilize multi-level (ML-SSM). In ML-SSM, whole divided into patches, principal component analysis applied each patches. avoid inconsistency among introduce new constraint called adhesiveness overlap regions experiments, demonstrate improved by using introduced ML-SSM.