are you sure there's no left/right swap in either package? That is, do FreeSurfer and VBM agree on which hemisphere is left?
On Thu, 2 Jun 2005, Akemi Tomoda wrote:
Dear FS list,
Here I have the FreeSurfer and the VBM data for the 26 young adults between the different parameters.
I see that there is a Pearson's correlation of 0.9148 for the Left Hemisphere gmv between FS and VBM. For the right hemisphere there is a correlation of 0.9061. But when I compare the differences between the left and right hemispheres by FS and by VBM, I see that we get a negative correlation of -0.3806, p = 0.0551. And when we compare the laterality indices between the hemispheres for the 26 subjects from FS and VBM we get a correlation of -0.3464, p=0.0830.
The mean (N=26) Left Hemisphere VBM data = 337820 (mm–3) and for the Right Hemisphere it is 337020. With FreeSurfer we get Left Hemisphere = 363113 and for the Right Hemisphere = 368816. That is, with VBM the mean Left Hemisphere is larger than the right while with VBM the right hemisphere is the larger.
The question for me is why is there a discrepancy on the differences between the hemispheres? Does this mean that FS and VBM are giving different data in spite of the strong correlations when comparing the individual hemispheres? This seems so to me. How do we account for this discrepancy in the face of the strong correlations between the FS and VBM results for the same hemisphere?
Any suggestion or comments regarding this matter would be highly appreciated.
Best,
Akemi
Developmental Biopsychiatry Research Program McLean Hospital, Belmont MA
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer