Dear Jesper,
To add the answer to your first question:
Here is the mean for SSEs of distorted versus corrected:
fslstats dti_distorted_sse -m: 0.438754
fslstats dti_distorted_sse -M: 2.217524 // take only nonzero voxels
fslstats dti_corrected_sse -m: 0.326887
fslstats dti_corrected_sse -M: 1.658303 // take only nonzero voxels
Indeed, the mean does get smaller. However, I was expecting that a subtraction of sse_corrected from sse_distorted would give a relevant map that shows residuals specific to eddy distortions. Is this not the way to go?
Best,
--
Maria
|