dear FSL people,
we have a question regarding seed-based correlation analyses of resting state fMRI data
we performed a seed-target correlation analysis on resting-state data in the same way explained in O'Reilly et al. (2009 Cerebral Cortex), that is, regressing the time course (TC) of each voxel in the target brain region against the eigen TC of each seed region, also including in the model the eigen TC of each other seed region (7 in total), as well as their first derivatives, as confounds. We also included as additional confounds the following: motion parameters, global, CSF and white matter signal, as well as their first derivatives, and we bandpass filtered the data in the range 0.009..0.08 Hz
We would like to have your opinion about one issue that came to our mind: is it possible that the correlation values between seed and target voxels are influenced by their respective slice scan time? If so, to what extent can it be caused by subsampling/aliasing issues?
thanks a lot!
best regards,
filippo migliorati and leonardo cerliani
|