Not that I have seen. Anytime people compare voxel extents, I am very
cautious. You can't use extent from group maps to conclude that they
are different. For example, if you have 50 voxels in test1 and 60
voxels in test2 out of the 100 voxels tested. (1) Clusters could be in
different areas within the 100 voxels so extent doesn't tell you
anything; (2) assuming that the 50 voxels are within 60 voxels, then
you still don't know if the the 10 additional voxels are different in
the amount of atrophy. Specifically, if the t-statistic is 3.01 in the
10 additional voxels in the significant group and 2.99 in the
non-significant group, then the tests aren't different. Furthermore,
the t-statistics doesn't tell you the amount of atrophy as the
variance of the subjects could be different in the two tests.
If possible, test for an interaction in a larger model.
Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Postdoctoral Research Fellow, GRECC, Bedford VA
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.
On Thu, Jul 26, 2012 at 9:13 AM, Dr. Juergen Dukart <[log in to unmask]> wrote:
> Dear SPMers,
>
> is anybody aware of if there is any valid (if possible published) way of comparing cluster extents of significant atrophy (as measured by VBM) from two different analyses (with the same degrees of freedom)?
>
> For the statisticians: Would it be valid way to apply a proportions test comparing two proportions calculated as the amount of "activated" voxels devided by the total amount of tested voxels and to use the following information provided by the SPM stats to calculate the "sample sizes": "total number of resels"/("expected cluster extent in resels" * "number of expected resels")? The denominator would then scale the total number of resels by the total expected extent of activated resels at a given statistical threshold.
>
> Thanks,
>
> Juergen
|