Dear Gang,
I have several questions about 1dGC. I hope you could help me.
I have an fMRI run with 3 task conditions, in block design, and I have baseline resting period before the first task block. And I’d like to calculate causality for each condition separately.
1. Specifying baseline regressor and task block of no interest in the covariates?
In the 1dSVAR page, it wrote: “It is the change relative to the baseline that is comparable across blocks/runs, regions, and subjects, therefore signal normalization through scaling in terms of the loose concept of percent signal change is very important, and it can be done during the pre-processing, or you can leave it for 1dGC.R to handle.” But how could I tell the program when was the baseline period? or shall I normalize the data manually before putting them into the model? About the regressors for tasks of no interest, I have extracted those columns from the design matrix. Shall I also create a baseline regressor in the same way?
2. ICA-cleaned time-series
I have my data cleaned using ICA decomposition in FSL. I’d like to use time-series extracted from the cleaned data, as I suppose that removing structural noise could improve the result. However, FSL performed temporal filtering before ICA, which I know it was not recommended. What do you think about this step?
3. White-matter/CSF/Global signal
Do you recommend to include those values as covariates?
Best regards,
Clive