Thanks to all of you for ideas. You may also find interesting to look at:
page on VBM, which has large section of Criticism. The most striking for me is the inaccuracy of the spatial normalization, which may account for as much as 10.8 mm.
On Thursday 15 May 2008 18:58:15 rahul@nmr.mgh.harvard.edu wrote:
Hi Martin,
Not to belabor this topic but one of the major advantages in using Freesurfer is that for an individual subject, you can ascertain what the individual volume/thickness is of specific neuroanatomic areas. Clinically, this information is of significant importance as you can then detail whether atrophy in specific regions is predictive of a clinical endpoint (e.g diagnosis of Alzheimer's disease) or how atrophy in specific regions progresses over time (e.g. longitudinal anatomic change within a specific individual).
Indeed that is a strong argument! However, to be able to say that "this individual is likely to have Alz due to his parietal cortex atrophy" one either needs to know exactly what the thickness should be at the age of the individual, or have several age matched controls in hand.
To the best of my knowledge, SPM/VBM only allows for group comparisons, not indidividual subject morphometric information, and
I am not an expert in VBM, but I wonder, if I would have those several controls in hand, shouldn't I be able to compare the individual with this group?
Thanks,
Martin