LOGARITHMIC PRIORS FOR LEVEL SET BASED SEGMENTATION OF DEFORMATION
We proposed the Large Deformation Unbiased Diffeomorphic Nonlinear Image Registration model, which has been successfully used to obtain theoretically and intuitively correct deformation maps. In this work, we extended this idea to simultaneously registering and tracking deforming objects in a sequence of two or more images. We generalize a level set based Chan-Vese multiphase segmentation model to consider Jacobian fields while segmenting regions of growth and shrinkage in deformations. Deforming objects are thus classified based on magnitude of homogeneous deformation.
| template image | study image | |||
 |
 |
|||
| deformed template | segmentation
of deformed region |
|||
 |
 |
|||
|
Segmentation of deformation for 3D serial MRI image. Volume cuts of template, study, and deformed template images are shown. The surface (zero level set) of the ventricle is shown. The ventricle surface is enlarged for better visualization. |
||||
| THREE-DIMENSIONAL SERIAL MRI VOLUMES | ||||
| template image | ||||
 |
 |
 |
||
| study image | ||||
 |
 |
 |
||
| deformed image superimposed with the segmentation of deformation | ||||
 |
 |
 |
||
|
axial plane |
coronal plane |
sagittal plane |
||
|
|
|
|
|
|
References:
Igor Yanovsky, Paul Thompson, Stanley Osher, Luminita Vese, Alex Leow, Multiphase Segmentation of Deformation using Logarithmic Priors, IEEE Computer Society Workshop on Image Registration and Fusion, pp. 1-6, June 2007.
