Applying the Image Registration Transform to Other Images

You can estimate the transform that maps an image onto the fixed image, and then apply the same transform to second image. This might be useful where the second image has much poorer signal-to-noise ratio, or different contrast that makes it difficult to estimate the transform. Application of a previously-found correction is also very quick, so this could save considerable time in cases where you know that the transform is the same for a series of images. Finally, you may want to make sure that the same transform is applied to a series of images.

Saving the Transform to Disk

To save the estimated transform to disk, ensure that the Select to save the registration transform to disk check-box is ticked. When you perform the registration, an additional file will now be created for each image to register: this will have an extension ".rtp" and will contain the registration transform for that image.

Applying the Transform to Another Image

To apply the transform to another image, ensure that the Select to read a
  registration transform to disk check-box is ticked, and select the ".rtp" file that contains the transform you want to apply:

Selecting the registration transform file

The same transform will be applied to all the images to be registered.

If you want to invert the transform, before you apply it, select Select to invert the registration transform before applying it check-box. You may want to do this if you want to register what was previously the fixed image to what was previously a moving image.

Registering using Image Header Information

Image position information may have been set by your image collection device (such as an MRI or CT scanner) and may be present in the image header if you have done all the image conversion and processing using Jim. If so, then two images can be brought into alignment using the header information along by selecting Select to register using image header information. When selected, the image to be registered is simply resampled at the pixel positions of the fixed image, using the header positional information to calculate the pixel intensities in the registered image.
Jim Home