Log-unbiased elastic image registration with spatial constraint for 3D CT lung images

Registering diagnostic lung CT and whole body CT images is a difficult task due to their acquisition under different breathing stages. We have implemented a novel framework for 3D CT lung image registration which combines elastic registration with log-unbiased deformations and a spatially variable constraint to reduce image folding and retain the rigidity of the bones. A comparison of the proposed method, versus classic elastic registration on 3D phantom data, has shown that our algorithm has been successful in keeping the ribs and other bony structures rigid while reducing the amount of folding of the deformation field.