Alignment by Maximization of Mutual Information
International Journal of Computer Vision
Interpolation artefacts in mutual information-based image registration
Computer Vision and Image Understanding - Special issue on analysis of volumetric image
Variational Methods for Multimodal Image Matching
International Journal of Computer Vision
A Viscous Fluid Model for Multimodal Non-rigid Image Registration Using Mutual Information
MICCAI '02 Proceedings of the 5th International Conference on Medical Image Computing and Computer-Assisted Intervention-Part II
Fast Fluid Registration of Medical Images
VBC '96 Proceedings of the 4th International Conference on Visualization in Biomedical Computing
Deformable templates using large deformation kinematics
IEEE Transactions on Image Processing
Hi-index | 0.00 |
We propose the representative slice method. This method is a parameter adjustment technique for viscous fluid registration. We experiment a whole-body CT image and whole-body MRA image registration using this method. The representative slice method is mainly consisted in three steps. In the first step, we divide whole-body image into six segments (head, chest, abdomen, hips, legs, and feet). In the second step, we select representative slices from six segments. In the last step, we regulate the viscosity for viscous fluid registration from 13 values of the viscosity. The selective viscosity is provided the well alignment in each segment's slices. In this way, the parameter regulation is performed for only the representative slices. Using this method, we can save a lot of time to regulate the viscosity and other parameters in whole-body image registration. After whole-body registration, we find some misregistered area. So we define these areas as new segments, and regulate the viscosity using the representative slice method. As a result, only 109 in 1,687 slices cannot be aligned. Thus we can register almost all whole-body CT image and MRA image, and the same time, we can also register two three-dimensional human atlas images using same displacement.