External Email - Use Caution        

Dear Freesurfer experts,

Sorry to post this issue again, but I'm not sure how to move forward with my data. I'd really appreciate any answer, because I'm under a lot of pressure. Using FS 7.1.0 I found inaccuracy of the WM surface near the putamen and accumbens regions. WM surface of the ctx-rh-medialorbitofrontal goes into the right putamen/accumbens region. Douglas replied that it's a bug in version 7.[012] and this area is probably not valid. Unfortunately, I've results in the affected region (ctx-rh-medialorbitofrontal) and this type of WM inaccuracy is quite common (~each fifth subject).  Both the putmen/accumbens regions and  the ctx-rh-medialorbitofrontal looks quite good in the aparc.DKTatlas+aseg.mgz (although WM goes into the putamen, in the aparc.DKTatlas+aseg.mgz the affected part is correctly marked as putamen and not as a cortical region, so it looks good). However, I don't know whether this WM inaccuracy affects statistics (aseg.stats and *.aparc.DKTatlas.stats) or not. I also did vertex-wise analysis, that gave significant result in the  ctx-rh-medialorbitofrontal, but I don't know whether the affected WM part has an effect on the results or the software realizes this WM error based on the results of subcortical segmentations (aseg) and corrects it somehow or excludes this part from the analysis.
Is there any suggestion how to move forward with this issue to be sure that our results are valid?
- Are my aseg.stats data valid for the putamen if WM goes into the putamen, but the putamen segmentation looks nice based on aseg.mgz?
- Are my  *.aparc.DKTatlas.stats data valid for the ctx-rh-medialorbitofrontal if it's WM border goes into the putamen, but the  ctx-rh-medialorbitofrontalsegmentation looks nice based on aparc.DKTatlas+aseg.mgz?
- Are my vertex-wise analysis valid for the region of ctx-rh-medialorbitofrontal?
- Is there any way to correct this type of WM surface somehow in FS7.1.0?
- It is not intended, because we already did a lot of corrections, but as a last resort, is it worth to rerun everything using an earlier version of FS?
Thank you.
Gabor




Gabor Perlaki <petzinger.gabor@gmail.com> ezt írta (időpont: 2022. márc. 28., H, 14:55):
Can you please help me with which measures are not valid in this case? Are the cortical (e.g. thickness of ctx-rh-medialorbitofrontal) and subcortical measures (e.g. right pallidum) both invalid? aseg.mgz looks nice and aseg+aparc.mgz looks also nice. Only the pial seems to be affected.
Thank you in advance for your help.

Gabor Perlaki <petzinger.gabor@gmail.com> ezt írta (időpont: 2022. márc. 26., Szo, 12:56):
Could anybody please help me with this?

Gabor Perlaki <petzinger.gabor@gmail.com> ezt írta (időpont: 2022. márc. 22., K, 0:31):
Dear all,

fswiki suggests that " There are regions where the surfaces are not intended to be accurate that you should be aware of: Areas around the hippocampus and amygdala. "

I also found inaccuracy of the WM surface near the putamen and accumbens regions. WM surface goes into the putamen (please see the attached picture). Do I need to fix this error or is it normal? Are the thickness of ctx-rh-medialorbitofrontal and the volume of right putamen valid in this case?
Thank you in advance,
Gabor



--
Gabor Perlaki
research associate
Diagnostic Center of Pécs
H-7623 Pécs, Rét str. 2.
Tel.: 0036-30-2084367
E-mail: petzinger.gabor@gmail.com


--
Gabor Perlaki
research associate
Diagnostic Center of Pécs
H-7623 Pécs, Rét str. 2.
Tel.: 0036-30-2084367
E-mail: petzinger.gabor@gmail.com


--
Gabor Perlaki
research associate
Diagnostic Center of Pécs
H-7623 Pécs, Rét str. 2.
Tel.: 0036-30-2084367
E-mail: petzinger.gabor@gmail.com