> I am coregistering 2 MRI.img brains. I have a mask that needs to remain in
> alignment with the source image, which I select in the "Other Images".
> The coresigtration of the source brain works fine. However the registered
> mask output is currpoted (either totally blank (for binary mask, or
> trancuated for values ranging from 16000-32500). I though it had to do with
> uint16/ little Indian, but it keeps happening for any normalization of the
> mask intensity values.
What kind of interpolation are you using for the reslicing. Nearest neighbour
is more possibly appropriate for binary data? What scalefactors do you use,
and what are the original values in the data?
Best regards,
-John
|