Lots of questions about lesions:
1) What is the best way to measure the size of a lesion in Freesurfer? Or should it be drawn in afni and transported to Freesurfer somehow?
2) What is the best way to measure how much of each ROI is affected by the lesion?
3) Line 25 lists "Left-lesion" and Line 57 lists "Right-lesion" in aseg.stats, but these are currently all zeros in brains that have clear lesions -- why? And why are those lines there if Freesurfer isn't recognizing a hole in the head as a lesion? Is there some option that must be selected to make this work?
4) Is it true that cortical lesions can throw off Freesurfer's ability to see landmarks, so that not only the lesion area might be mislabeled, but surrounding tissue and even distant tissue too? If so, is there a way around this problem built into Freesurfer? I have one brain with a large cortical lesion that looks like it parcellated okay, and another with a huge lesion that doesn't look right at all. Periventricular lesions look fine -- is this what others have found?
5) What is the best way to dump the parcellation data and the subcortical segmentation data back into afni (which files specifically) so that it can be viewed in 3D on the original structural brain as a mask?
Help would be much appreciated. Many thanks, Joan Fisher University of Chicago
Hi Joan,
1-3) We haven't done a whole lot of lesion quantification. We've done some more recently, but it needs registered T2/PD scans, and isn't distributed yet. So I guess drawing in AFNI is the easiest thing to do at the moment. The lesions are more commonly labeled as "white matter hyperintensity", but hey are pretty ambiguous without T2 and PD info.
4. I haven't seen this happen for cortical lesions. If the white matter surfaces of the recons are accurate, then the parcellation should be fine. Can you send us an example of one that failed?
5. I think if you have a surface loaded in tkmedit you can import the annotation, but maybe Doug and Kevin can comment?
cheers, Bruce
On Fri, 12 Jan 2007, Joan Fisher wrote:
Lots of questions about lesions:
- What is the best way to measure the size of a lesion in Freesurfer? Or
should it be drawn in afni and transported to Freesurfer somehow?
- What is the best way to measure how much of each ROI is affected by the
lesion?
- Line 25 lists "Left-lesion" and Line 57 lists "Right-lesion" in
aseg.stats, but these are currently all zeros in brains that have clear lesions -- why? And why are those lines there if Freesurfer isn't recognizing a hole in the head as a lesion? Is there some option that must be selected to make this work?
- Is it true that cortical lesions can throw off Freesurfer's ability to see
landmarks, so that not only the lesion area might be mislabeled, but surrounding tissue and even distant tissue too? If so, is there a way around this problem built into Freesurfer? I have one brain with a large cortical lesion that looks like it parcellated okay, and another with a huge lesion that doesn't look right at all. Periventricular lesions look fine -- is this what others have found?
- What is the best way to dump the parcellation data and the subcortical
segmentation data back into afni (which files specifically) so that it can be viewed in 3D on the original structural brain as a mask?
Help would be much appreciated. Many thanks, Joan Fisher University of Chicago
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Joan Fisher