Skip to main content
. 2010 Jun 9;11(3):101–123. doi: 10.1120/jacmp.v11i3.3175

Figure 4.

Figure 4

Axial slices of colored source images overlaid on greyscale target phantom images for two selected registrations: rigid alignment of CT images (a) in which a single target has been translated 8 mm; CT images after deformable registration on Reveal‐MVS (b) demonstrating that the system does not move the sphere back to its original location; rigid alignment of MRI and CT images (c) in which a sphere in the MRI has approximately doubled in volume. An inset window with 5x magnification depicts the differences between the phantom perimeter in the MRI and CT images resulting from a lack of MRI signal from the plastic phantom perimeter. MRI and CT images (d) after deformable registration in which the system attempts to deform the MRI phantom volume to that of the CT phantom volume while ignoring the volumetric differences in the modified target.