Hi Kristine
the aseg.stats volume uses partial volume correction to get subvoxel accuracy.
cheers Bruce On Fri, 23 Mar 2018, Lee Subin Kristine wrote:
Hi,
I have a question about volume info from freesurfer vs matlab.
I am looking at, say, the left hippocampus volume from asegstats, which was 3778.2 mm^3. When I try to get the same volume by loading aseg.nii on matlab, however, I get a different number, 3945.
For matlab, I used the following command, which counts the number of voxels equal to the left hippocampus segment id: length(find(aseg.img==17))
This was done after I converted the original aseg.mgz file into native space, using 2 different methods: (1) mri_convert -rl rawavg.mgz -rt nearest -i aseg.mgz -o bert_aseg.nii
(2) mri_label2vol --seg aseg.mgz --temp rawavg.mgz --o bert_aseg.nii --regheader aseg.mgz In both cases, the output was 3945.
The rawavg.mgz used here is also 1mm isovoxel, so I don't think it is due to the difference in voxel dimensions.
What could be the cause of the difference in volume?
Thanks,
Kristine
Sent from Mail for Windows 10