Hi Alan,
in most cases, the segmentation process should be identical. Most parameter settings should apply to all datasets with a reasonable contrast between grey and white matter. However, you should include "site" as an additional factor in your statistical analysis. If you plan to analyse data of different populations (e.g. patients and healthy controls) you should acquire all populations on all scanners. If you acquire e.g. patients on scanner A and controls on scanner B you may not be able to distinguish between scanner and population effects.
Hope this helps
Volkmar
|