Hi,
On 19 Mar 2009, at 12:59, Raymond Salvador wrote:
> Dear FSL list members
>
> I've been trying to use the cluster tool to obtain cluster corrected
> output images from a z-stat map (generated by some self written
> code). However, I have some questions:
>
> 1.- I want to make "one tailed" tests (only considering positive z-
> values, as they originally come from an f-map through an f-to-p-to-z
> transformation). Can I just set the threshold probability (-p
> parameter) to "2*0.05" and consider only those significant clusters
> with positive-z values in the output?
The cluster p-value stuff already assumes one-tailed testing, I believe.
>
> 2.- For the -d parameter [sqrt(det(Lambda))] I have made a very
> coarse calculation (and probably quite wrong) taking, as starting
> point, the value of the standard deviation of a gaussian spatial
> filter I applied to the individual images.
>
> Specifically, considering sigma = 3 mm, i ended up with a
> sqrt(det(Lambda)) = 0.013 by a simplified formula
>
> sqrt(det(Lambda)) = 1 / (2*sigma^2)^(3/2)
>
> It this any good? Otherwise, if I want to take a conservative value
> for sqrt(det(Lambda)), which value could I apply?
You should just be able to use the "smoothest" program to estimate
this all for you....for more theory info see our old technical report
on smoothness estimation by MJ.
>
> 3.- Finally, when i run the tool with the following line
>
> cluster -i inputzimage -t 2.3 --volume 169115 -p 0.1 -d 0.013 --
> othresh ouputimage
>
> I keep on having the following complain:
>
> --volume: Missing non-optional argument!
For any options with two dashes at the front you have to use "=" and
no spaces, like --grot=grotty
Cheers.
>
> Well, I hope my questions aren't too silly.
>
> Thanks a lot in advance,
>
> Raymond Salvador
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|