Brain Follow-Up - Processing Options

Having set up the Brain Follow-Up tool's input images as detailed in the Introduction, you can now set up the other processing options:

• . Only available when there are more than two time-points. Set the number of iterations to be performed, between 1 and 3. This affects the type of processing performed and the time required to do it.
  • During the first iteration, the registration target image is the template image, which is either a built-in template or the one you set with the option below. All images with the first image contrast are registered to this target.

  • If you set at least 2 iterations, a registration target is then formed from the images just registered, to form a patient-specific registration target. The patient-specific registration target is in a spatial position that is the geometric average position of all the input images. This target is used during iteration number 2.

  • If you set 3 iterations, a second, more detailed patient-specific registration target is formed from the images just registered during iteration 2. The result is a more precise patient-specific registration.

  Each extra iteration results in a more accurate inter-time-point registration, and therefore better subtraction images. However, each extra iteration takes a while to perform, particularly if you are analysing a large number of time points. In general you are recommended to use 3 iterations unless you only require a basic registration.

• . Select this option if you want to use Brain Finder to isolate the brain as part of the analysis. This can help in two ways:
  1. Isolating the brain can provide a cleaner registration target, without the surrounding tissues (skin and subcutaneous tissues) that are sometimes distorted.
  2. When reviewing the images, is is often easier to concentrate on changes within the brain if the surrounding tissues are not present.
  If you choose this option, you will also need to select the type of contrast in the first input image, from T1 (T₁-weighted), T2 (T₂-weighted), or PD (Proton Density).

• . Select this option if you want to use the brain mask (found in the step above) to mask the brain in the output difference images. This option is only available if you chose to use the Brain Finder with the option above.

• . Select this option if you want the Brain Follow-Up tool to calculate the magnitude (absolute value) of the output difference images, rather than the signed difference. This can be useful if you are expecting to see, and assess, both negative and positive difference, because when displaying the signed difference, negative differences can often be missed (because they are displayed darker than the background).

• . Only available when there are more than two time-points. Select this option to tell the Brain Follow-Up tool that your image slices are in reverse order compared to the template image. Jim's built-in head template image has slices that are in inferior-to-superior order with increasing slice number. If your image slices are in superior-to-inferior order, then select this option, which will reverse the order in the template.

• . Only available when there are more than two time-points. Jim has a built-in registration target (template) image used in the first stage of registration. If this does not work well, and you want to try using your own registration target image, then select this option and select that target image using the image selection panel below.

• . If you select this option, Jim will write the final registration target, used in the final registration iteration to a disk image. The name will be created from the base image name you set with the suffix "Template". This can be useful if you want to verify that the final target image is sensible.

• . If you select this option, Jim will write the final registration transforms found. These can later be used by the Image Registration tool and applied to other images, if you want to transform them in the same way.

• DM_N. This is the difference image between time point M and time point N. This is the example above, this would be the difference image between time point 1 and time point 3. All possible combinations of difference are created. For example, if you analyse 3 time points, the Brain Follow-Up tool will create D1_2, D1_3 and D2_3 difference images.
• CX. This is for image contrast X. There will be one set of difference images for each image contrast. If you use only one image contrast, then C1 is not appended.