3D Medical images registration

Image registration is the process of overlaying images of the same scene. As we search for the best alignment of the two images by transforming one into the other, it is a very crucial issue to assess how similar two images actually are. The next step in computer aided diagnosis is called “assessment of the similarity”. There are two main classes of similarity measures namely feature-based and intensity-based. In this paper two similarity measures: NCC-normalized cross correlation and GD-gradient difference are mentioned. Normalized Cross Correlation has been used for various registration problems, because difference in contrast and brightness should not affect the similarity measure. Goal the current study is to align T1-weighted and T2-weighted MR knee slices. Both sequences are converted to a fuzzy representation. Then, the entropy and energy measures are employed in the NCC and GD methods. The alignment based on energy and entropy fuzzy mcasures shows a significant improvement in comparision with thc implementation of the original image.