Use the "surface-based-volume". As the name implies, the surface-based uses the surface, which has subvoxel accuracy. The Left-Cerebral-White-Matter is based on counting 1mm voxels.

doug

saulg@ualberta.ca wrote:

Hi there,

In the aseg.stats file, there are two reports of cerebral white matter volume. The first report is near the top of the file, as such:

# Title Segmentation Statistics # ... # surface-based-volume mm3 lh-cerebral-white-matter 172917.046875 # surface-based-volume mm3 rh-cerebral-white-matter 173269.906250 # surface-based-volume mm3 tot-cerebral-white-matter 346186.953125 ...

The second report is lower in the file, contained in the section that would be turned into a spreadsheet by the 'asegstats2table' function:

# ColHeaders Index SegId NVoxels Volume_mm3 StructName normMean normStdDev normMin normMax normRange 1 2 167965 167965.0 Left-Cerebral-White-Matter 104.1714 8.0979 41.0000 131.0000 90.0000 ... 21 41 179570 179570.0 Right-Cerebral-White-Matter 102.5936 8.2544 50.0000 133.0000 83.0000 ...

I noticed these numbers are consistently different, so I have two questions:

1. Why are they different? Is it because the first is surface based?
2. Which numbers should I use when reporting total cerebral white

matter volume?

Thanks so much!

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer