Paper Info
Title
PHYSICALLY ACCURATE B-SPLINE BASED NON-RIGID REGISTRATION USING VARIABLE SPRING MODEL
Abstract
This paper presents a physically accurate fast parametric non-rigid registration method based on B-spline deformation model. Our method is based on using variable spring model to add a physical constraint to ensure that the deformation does not violate elasticity characteristic of the organs. Using the assumption of human organs as elastic bodies, we modeled the whole volume as 3-D springs connecting each neighboring voxels. Registration was performed on B-spline deformation model with the combination of SSD and total spring energy as optimization criterion. We also added the concept of elasticity map to made it possible to model a more specific elasticity characteristic such as hardness/softness of the organ. We showed that compared to the original algorithm, our algorithm was able to realize a more realistic deformation. We also applied our method to register 3D CT images of human heart from different heart beat phases.
Year
DOI
Venue
2007
10.1109/ISBI.2007.356960
Arlington, VA
Keywords
Field
DocType
biomedical imaging,cardiology,computerised tomography,image registration,medical image processing,3D CT images,B-spline deformation model,SSD,elasticity map,human heart,image registration,total spring energy,variable spring model
B-spline,Voxel,Spline (mathematics),Computer vision,Computer science,Medical imaging,Parametric statistics,Artificial intelligence,Deformation (mechanics),Elasticity (economics),Image registration
Conference
ISSN
ISBN
Citations 
1945-7928
1-4244-0672-2
0
PageRank 
References 
Authors
0.34
6
5
Name
Order
Citations
PageRank
Nicholas Herlambang100.68
Hongen Liao239070.91
Kiyoshi Matsumiya3266.28
Ken Masamune425348.57
Takeyoshi Dohi549697.71