Warping T1s with lesions to a base template space

In a previous post, the recommended steps to warp brains with lesions were this:

[li] Draw the exclusion mask over the subject’s image dataset.
[/li][li] Register the template to the subject’s dataset, and then apply the inverse of the warp transformation to get the subject’s exclusion mask into template space. We need to do this because ‘-emask’ option applies to the base dataset.
[/li][li] Manually assemble correct processing stream using the exclusion mask. Use @SSwarper or auto_warp.py as a reference for the processing stream.

For 2, what should be used for registration? Should I use align_epi_anat.py?

For 3, I have not looked into it too much right now, but I plan on using @SSwarper as a reference to create the warping stream. I am assuming I will use the ‘-emask’ option for all calls to 3dAllineate and 3dQwarp.