Hello,
I examined data from a public database (ABIDE) processed both with freesurfer 5.1 and freesurfer 6.0 and I noticed that in version 5.1 the equality "SubCortGrayVol"+"CortexVol" ="TotalGrayVol" holds for every patient while it is never true in version 6.1.
I attach as proof two histograms of the distribution of the quantities in the equation above calculated on a dataset of 1076 patients.
[cid:d80a9491-4bf9-4d0c-b4d8-1c62f6e13aab]
For clarity I also attach the aseg.stats files of one patient obtained with the two different freesurfer versions in which I have highlighted the data I'm talking about.
[cid:3eb0d89a-ff21-40c8-b3b1-d3ab0b43392a][cid:54bbdf12-ce54-4e8a-9eee-ec9d43f0e600]
I observed that for almost every patient the values of "TotalGrayVol" and "CortexVol" haven't changed between the two versions while "SubCortGrayVol" differs widely.
[cid:d0af9a9d-e9a3-4910-a0aa-03dd0874d4b1]
I am thus asking if in version 6.0 "TotalGrayVol" is obtained from the sum of more than two quantities and not anymore by "SubCortGrayVol"+"CortexVol", if yes what are the quantities involved? Or could this be just due to an error in the measurement of "SubCortGrayVol"?
I also have noticed that in files lh and rh aparc.stats (of FS 6.0) sometimes there are less cortical structures than the usual 34. I have always thought that this was due to the fact that freesurfer didn't manage to define these regions, however looking at aparc.DKTatlas.stats (FS 6.0) I realized that sometimes there are more than the usual 31 cortical structure.
In particular sometimes appears the temporal pole, that doesn’t belong to the DKT atlas. What could be causing this error?
I attach an example file.
[cid:5c1e1627-6cd4-47e0-813d-1da5ca283470]
Thank you for your attention.
Elisa
