ABSTRACT

It is common for patients to undergo multiple tomographic radiological imaging for the purpose of medical diagnosis. These multi-modality images provide complementary information. However, it is difficult for doctors to fuse these images exactly because of the variations in patient orientation. For this reason, there has been considerable interest in using image registration techniques to transfer all the information into a common coordinate frame. The objective of this study is to find a transformation that best aligns the float image (unregistered image) with the reference image.