We have done some similar alignment and skullstripping. Because rat brains are imaged in many different ways, it's difficult to know ahead of time which methods will work. 3dIntracranial and 3dAutomask may be useful alternatives to 3dSkullStrip. You can use the "-rat_align" option for align_epi_anat.py (which calls 3dAllineate). To compare similar types of images, the lpa or nmi cost functionals work best with the "-cost" option. In some cases, the "-edge" option helps align images that change in intensity.
