 |
 |
Hi everyone,
I want to apply DCM on resting state fmri. To extract LLF I used the method for design matrix as in this article.
http://www.ncbi.nlm.nih.gov/pubmed/23927904
Signal from VOI extract by eigen variate button on result of SPM.mat following using VOI mask
Using DCM tab specify the model then it asks for effect of eight conditions (boxcar functions) as in the paper. For all VOIs consider all boxcar as input. In the paper it is specified fourier sets consider as driving input for all vois. Is this the correct way of specify ing? I specified all possible models. But it provides fully connected VOIs as best model. Can anyone help me to solve this.
Alternatively I tried GLM DCT as in the following publication
Specified models using DCM. But it doesn't ask for effects of DCT. Extracts VOIs as previous. It also provides fully connected VOIs as best. I applied this DMN and the cortico strital network.
In the second paper the network is clearly defined , only exogenous fluctuations distinct the models . In the first article the direction of connectivity is derived using DCM. My doubt should we have specific basic model for VOIs? How can we assume endogenous fluctuations between connections in resting state.? Can anyone help me to sort it out?
Thanks in advance
Regards
Uma