On Wed, 15 Dec 2004, Mark Jenkinson wrote:
> Hi Martin,
> I hadn't forgotten your email, I just needed time to make a schedule
> file to do what you wanted.
I am very sorry if I have urged you, which certainly didn't mean to. I
understand that with FSL gaining more and more popularity you guys are
getting exceedingly busy with support. I am aware that my last
question was a bit unusual (and few last emails to the list too) and I am
glad that you found a time to answer. FSL is a fantastic tool and I count
myself lucky that I can use it.
> There are three things you need in the schedule file:
> measurecost (which you had) , printparams and setscale commands
> Flirt (at present) requires all the scales to be stepped through
> sequentially in order to generate the appropriate reference and input
> images internally. Hence you need setscale for 8 then 4 then 2 then 1.
> I'm attaching my schedule file (showcost) which printed the correct cost
> function measurement for me. You need to use it in registration
> mode, rather than applyxfm mode. For example:
> flirt -in epi -ref highres -schedule showcost -init epi2out.mat -omat
> where epi2out.mat is the previously generated registration matrix,
> and -omat /dev/null is just used to junk the printing out of the final
> matrix (which of course is just the same as the -init in this case).
Thanks for the schedule file I will try it later today and let you know how it works.
I am sure it will work just fine.
> Let me know if you have any trouble with this.
> All the best,