Dear Cornelius,
FSL doesn't provided uncorrected cluster P-values anywhere, and there's no convient way to generate them for conjunction cluster inference either.
If it helps, if all you want is uncorrected conjunction P-values, you don't need the easythresh_conj script. Voxelwise conjunction inference simply consists of taking the minimum of a series of statistic images, and then assessing that image as if it were a single image.
Hence, you could just use fslmaths; in a Feat directory...
fslmaths stats/zstat1 -min stats/zstat2 -ztop stats/conj_12_uncorrp
fslmaths stats/conj_12_uncorrp -mul -1 -add 1 -thr 0.999 stats/conj_12_1-uncorrp_thresh
will give you the uncorrected voxel-wise P-values for a conjunction of contrast 1 & 2, and a thresholded 1-P image for 0.001 uncorrected.
Hope this helps!
-Tom
On Thu, May 5, 2011 at 10:50 AM, Cornelius Werner
<[log in to unmask]> wrote:
Hello all,
I am trying to do a conjunction on two interaction contrasts. Effects
will not be huge. Therefore, I'd like to go for an exploratory
uncorrected p-value of 0.001 - is that possible? Within the script,
there is the cluster command doing the stats, obviously yielding
cluster corrected p's. Is there a workaround?
Thanks in advance,
Cornelius
--
Dr. med. Cornelius J. Werner
Department of Neurology
RWTH Aachen University
Pauwelsstr. 30
52074 Aachen
Germany
--
____________________________________________
Thomas Nichols, PhD
Principal Research Fellow, Head of Neuroimaging Statistics
Department of Statistics & Warwick Manufacturing Group
University of Warwick
Coventry CV4 7AL
United Kingdom
Email:
[log in to unmask]Phone, Stats: +44 24761 51086, WMG: +44 24761 50752
Fax: +44 24 7652 4532