[MINC-users] rawtominc & nu_correct difficulties

[Andrew Janke]

> it worked! This result is as good as anyone can expect. Can you explain to me the significance of the parameters, and why exactly it worked?

One of the at times amusing things about some of the addon MINC tools
(minctracc and nu_correct in particular) is that they were very much
optimised for "head sized" data. It is a common trick to make data
"head sized" to make things work at times.

You can make it work on "small data" but there are a number of
parameters that must be adjusted.  For example in minctracc you MUST
adjust -w_rotations for example.  I shall leave that one as an
exercise for the reader. :)

> Actually, I may have to go back and redo it with the actual voxel spacing, because the non-rigid registration with patient data, which comes afterward, must be applied to anatomical meshing... but everything scales accordingly, right?

in minctracc, yes almost everything does.  For nonlinear you should be
OK so long as you set the step correctly.


-- 
Andrew Janke   (a.janke at gmail.com || http://a.janke.googlepages.com/)
Canberra->Australia    +61 (402) 700 883