[Freesurfer] aseg.stats vs. volumes in native T1 space

Dear all,

I've run the freesurfer on 3 subjects and converted the aseg.mgz into *.nii.gz in the native space:

"mri_convert -rl rawavg.mgz -rt nearest aseg.mgz aseg2raw.nii.gz"

I've did a stats on left and right hippocampal volumes in the native space using fslstats:

"fslstats aseg2raw.nii.gz -l 16.5 -u 17.5 -V" for the left hippocampus

and

"fslstats aseg2raw.nii.gz -l 52.5 -u 53.5 -V" for the right hippocampus

The volumes given by fslstats was about 3% higher than the values reported in the aseg.stats file. The native resolution of the T1 images is 0.45*0.45*0.9 mm^3. I thought that observing a minimal difference could be the consequence of reslicing, but I really don't understand why fslstats gives consequently greater values (for both hippocampi in all three subjects).

Thanks, Gabor