Intermodality registration and fusion of liver images for medical diagnosis

SPECT is valuable for the evaluation of hepatic function and CT can offer better anatomical images for liver and its adjacent structures. Image registration from the two modalities, CT and SPECT, acquires both structural and functional information, and therefore is very useful for clinical diagnoses. The registration and fusion of CT and SPECT images are studied. We propose a two stage registration method based on principal axes transformation and surface fitting to register and fuse CT and SPECT images: First, the centroids and principal axes of these landmarks are calculated and aligned. Then the transformation matrix between CT and SPECT images is modified based on the least square error on the distance image. After registration, uniform volume data is reconstructed for image fusion.