Dear Bernie,
>and if so, how do I construct this? My guess is to use imcalc, but I have no
>idea of what operation to use. Could you give me a little bit more
information?
Yes, that exactly right. Use an expression like i1.*(i2 > 0) in imcalc
and apply it to the first [1st-level] contrast image to be analysed.
Now repeat the estimation. The voxels that are masked in the first
image will now be masked in all images and the search volume will be
restricted by i2 (the thresholded SPM{T}). Now test the second [2nd-level]
contrast with this anatomical restriction on the search space.
With very best wishes,
Karl
At 13:50 17/03/2005 +0000, you wrote:
>On Fri, 11 Mar 2005 13:23:04 +0000, Karl Friston
><[log in to unmask]> wrote:
>
>
>>If you wanted to pursue a more sensitive analysis you could try the
>>following procedure: At the second-(between subject)-level you have
>>two contrasts C1 and C2. Threshold C1 at p=0.001 (uncorrected) and
>>form a mask. Apply this mask to the images and threshold C2 at p=0.05
>>(corrected for the [masked] search volume). Save the resulting SPM{T},
>>which contains voxels that are significant for C2 at p=0.05 corrected.
>>Now repeat the procedure but swap C1 and C2. Voxels that are found in
>>both SPM{T} are significant at p=0.05 corrected in both contrasts and
>>constitute voxels where you can infer k = 2.
>>
>Dear Dr. Friston,
>
>I am a beginning SPM user, and fairly new to imaging. I have a similar
>albeit different problem as Uwe's, on which you commented.
>We have run an fMRI in which we want to look for conjunction areas for 4
>types of stimulus material in a comparison task (N=12). Applying a
>conjunction under the 'global null' gives nice results, in the sense that we
>find conjunction areas at an FDR corrected p < .05, in line with what we
>would expect from previous studies. However, under the 'conj null' nothing
>survives. Admittedly, one of the tasks gives below (corrected treshold
>activation) I guess. Nevertheless, I was very interested in the method you
>proposed to Uwe of masking the contrasts at different tresholds.
>
>My question now is how exactly to proceed. Let's say for 2 of my contrasts.
>More specifically, I managed to create an untresholded (.001) image for C1,
>but I do not understand what you mean with applying the mask to the images.
>Do you mean to construct a new image for C2, masked with this C1_treshold,
>and if so, how do I construct this? My guess is to use imcalc, but I have no
>idea of what operation to use. Could you give me a little bit more
information?
>
>Thank you very much in advance,
>Bernie Caessens
>Dept. of Experimental Psychology, Ghent University
>Ghent Institute for functional Magnetic Resonance Imaging, GIfMI
>
>
|