the registration of CT and MR images is important to analyze the effect of PCL and ACL deficiency on knee joint. Because CT and MR images have different limitations, we need register CT and MR images of knee joint and then build a model to do an analysis of the stress distribution on knee joint. In our project, we adopt image registration based on mutual information. In the knee joint images, the information about adipose, muscle and other soft tissue affects the registration accuracy. To eliminate the interference, we propose a combined preprocessing solution BEBDO, which consists of five steps, image blurring, image enhancement, image blurring, image edge detection and image outline preprocessing. We also designed the algorithm of image outline preprocessing. At the end of the paper, an experiment is done to compare the image registration results without the preprocessing and with the preprocessing. The results prove that the preprocessing can improve the image registration accuracy.
[1]
Max A. Viergever,et al.
Image registration by maximization of combined mutual information and gradient information
,
2000,
IEEE Transactions on Medical Imaging.
[2]
Bostjan Likar,et al.
A hierarchical approach to elastic registration based on mutual information
,
2001,
Image Vis. Comput..
[3]
George Lindfield,et al.
Numerical Methods Using MATLAB
,
1998
.
[4]
Paul A. Viola,et al.
Alignment by Maximization of Mutual Information
,
1997,
International Journal of Computer Vision.
[5]
Guy Marchal,et al.
Multimodality image registration by maximization of mutual information
,
1997,
IEEE Transactions on Medical Imaging.
[6]
Michael Brady,et al.
Phase mutual information as a similarity measure for registration
,
2005,
Medical Image Anal..
[7]
Rafael C. González,et al.
Digital image processing using MATLAB
,
2006
.