In fact, I create just a 4D image with the CTRL and PAT subjets and ran the comand line. I dont have idea how compare the maps obtained by seed analysis in the CTRl group and PAT group to find the diferences between groups.
RG
De: FSL - FMRIB's Software Library <[log in to unmask]> en nombre de Rosalia Dacosta Aguayo <[log in to unmask]>
Enviado: lunes, 8 de mayo de 2017 02:25 p. m.Dear Gabbo,
Just two simple questions....if you are doing seed analysis...what are you comparing with those contrasts?...did you used motion correction parameters as covariates and a white matter mask to regress out for de white matter signal? ....did you use FEAT?
El 8 may. 2017 21:08, "Gabbo Ramírez García" <[log in to unmask]> escribió:
These are the files of contrast and matrix
design.mat
/NumWaves 2
/NumPoints 12
/PPheights 1.000000e+00 1.000000e+00
/Matrix
1.000000e+00 0.000000e+00
1.000000e+00 0.000000e+00
1.000000e+00 0.000000e+00
1.000000e+00 0.000000e+00
1.000000e+00 0.000000e+00
1.000000e+00 0.000000e+00
0.000000e+00 1.000000e+00
0.000000e+00 1.000000e+00
0.000000e+00 1.000000e+00
0.000000e+00 1.000000e+00
0.000000e+00 1.000000e+00
0.000000e+00 1.000000e+00
design.con
/ContrastName1 CTRL>PAT/ContrastName2 PAT>CTRL/NumWaves 2/NumContrasts 2/PPheights 1.000000e+00 1.000000e+00/RequiredEffect 5.042 5.042
/Matrix1.000000e+00 -1.000000e+00-1.000000e+00 1.000000e+00
De: FSL - FMRIB's Software Library <[log in to unmask]> en nombre de Rosalia Dacosta Aguayo <[log in to unmask]>
Enviado: lunes, 8 de mayo de 2017 02:03 p. m.
Para: [log in to unmask]
Asunto: Re: [FSL] group analysisHi Gabbo,
It seems that something went wrong with your design. To see it, we need to see your design.mat and design.com.
Rosalia
El 8 may. 2017 20:59, "Gabbo Ramírez García" <[log in to unmask]> escribió:
The problem is that the image resulting is a image with just a 0.99 value, in fact is a wstrange result for that reason I guess that something is wrong.
I want to do Two-Sample Unpaired T-test, because I have a control group and a patients group. for that reason I did used tha command line.
I dont know if I am missing some step.
RG
De: FSL - FMRIB's Software Library <[log in to unmask]> en nombre de Blazej Baczkowski <[log in to unmask]>
Enviado: lunes, 8 de mayo de 2017 01:44 p. m.
Para: [log in to unmask]
Asunto: Re: [FSL] group analysisHello
It is difficult to guess what your problem is based on your description. What type of error do you get?
It seems that you want to run a two sample t test. For this you need to specify .grp file -- information on exchangeability, which is missing in your command line.
Best, Blazej
On 8 May 2017 at 20:11, RGuser <[log in to unmask]> wrote:
Hello FSLuser
I am doing seed-analysis for functional resting state images. I get the main signal of some seeds and correlate that signal with the global signal of the brain by Pearson correlation and the resulting image have p- values. I have two groups and now I want to compare the functional maps resulted after correlation signal seed - signal brain.
I used this line to do it but it did not work.
randomise -i 4dfuntionalmaps -o result -d design.mat -t design.con -m mask.nii.gz -n 500 -T
If somebody can help me I will be thankful.