Hi,
A few separate points here:
It is correct that the images in melodic_IC are divided by the noise
standard deviation images, and in the null case have a Gaussian
distribution.
I suspect that the reviewer doesn't have anything specific in mind
when they are stating that IC maps are not normally distributed......?
The most important issue though is that the within-map distribution is
not very relevant to whether a cross-subject test statistic is
normally distributed. If you do cross-subject voxelwise comparisons of
IC maps, then assuming that the method by which you chose the maps to
test is reasonable, then you _would_ expect the null distribution of
the voxelwise test statistics to be normal.
Cheers.
On 18 Dec 2008, at 16:36, Philipp G. Saemann wrote:
> Dear melodic experts,
>
> we are currently confronted with a reviewer's opinion that states,
> IC maps are
> per se not "normally distributed" and cannot undergo any voxelwise
> statistics.
>
> This we had done on Z-maps gained from individual ICAs (using the
> melodic_IC.img/hdr) after an automated component selection step, as
> e. g. in
> the work of Greicius et al. Resting-State Functional Connectivity in
> Major
> Depression: Abnormally Increased Contributions from Subgenual
> Cingulate
> Cortex and Thalamus, Biol Psychiatry, 2007; or in other work on
> clinical
> samples (e. g. Alzheimers disease, Sorg et al.)
>
> We understood from the description of FSL that the raw IC values are
> divided
> by the residual noise from the dissection into ICs, that results
> then in the Z
> maps.
>
> Can these Z maps can be subjected to voxelwise comparisons,
> typically using
> ANCOVAs (across subjects, e. g. 20 of a group A against 20 of a
> group B)?
>
> We are grateful to any comments on this issue -
> thank you very much in advance!
>
> best regards,
> Philipp
>
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
|