Dear Roohi,
have you tried using the DCT basis set I mentioned in an earlier email:
N = 230; % number of scans
TR = 2; % TR in {s}
h = [0.01 0.08]; % {Hz}
n = fix(2*(N*TR)*h + 1);
X = spm_dctmtx(N);
X = X(:,n(1):n(2));
save('DCT.txt','X','-ascii');
by including DCT.txt as 'Multiple regressors' and then using an
F-contrast over them.
Note that in the paper you are referring to they changed the default
high pass filter cutoff period of 128s to 100s.
Best regards,
Guillaume.
On 21/02/15 15:03, Roohi Jafari wrote:
> Dear Helmut and Guillaume
> Thank you very much. I defined my GLM through regressors and added head motion parameters to it, but still my results don't look like the paper (I tried it on 24 subjects but results are far from DMN). Now my GLM is completely similar to the paper, so I think there might be something wrong about my contrast. In paper it's been said that 'a diagonal F-contrast of all the eight regressors' so I defined my f-contrast for 8 regressors and 6 head motion parameters this way:
>
> 7/8 -1/8 -1/8 -1/8 -1/8 -1/8 -1/8 -1/8
> -1/8 7/8 -1/8 -1/8 -1/8 -1/8 -1/8 -1/8
> -1/8 -1/8 7/8 -1/8 -1/8 -1/8 -1/8 -1/8
> -1/8 -1/8 -1/8 7/8 -1/8 -1/8 -1/8 -1/8
> -1/8 -1/8 -1/8 -1/8 7/8 -1/8 -1/8 -1/8
> -1/8 -1/8 -1/8 -1/8 -1/8 7/8 -1/8 -1/8
> -1/8 -1/8 -1/8 -1/8 -1/8 -1/8 7/8 -1/8
> -1/8 -1/8 -1/8 -1/8 -1/8 -1/8 -1/8 7/8
>
> Is there anything wrong about it? I also tried eye(8)T but still no good results.
>
> Would anyone help me? I need to get some results about GLM analysis similar to this paper:
> http://www.ncbi.nlm.nih.gov/pubmed/?term=Identifying+the+default+mode+network+structure+using+dynamic+causal
>
> Thanks
> Roohi
>
--
Guillaume Flandin, PhD
Wellcome Trust Centre for Neuroimaging
University College London
12 Queen Square
London WC1N 3BG
|