Dear all,
(posting mail again, since my first mail seems didn't get posted...! sorry if it bothered you!)
I am trying to get a beta map(GLM) in CBV data using SPM12.
My data are acquired in 4 different conditions and I wanted to get a beta map of contrast between conditioned versus non-conditioned state using parametric modulation. I also added temporal derivatives and dispersion derivatives of the basis functions as regressors. By adding all options together, it gives a 6 regressors of interest. First 3 regressors are "onsets(+duration) of 4 conditions convolved with canonical hrf", "temporal derivatives of 1st regressor", and "dispersion derivatives of 1st regressor". The next 3 regressors make me little confused. 4th regressor is for parametric modulation which seems okay, but the other two(5th and 6th) are the temporal and dispersion derivative of the 4th regressor. I heard that using temporal and dispersion derivative is a way to model latency or duration of the peak response. I think this makes sense in the canonical hrf(first 3 regressors), but I am not sure this also makes sense in the regressor for parametric modulation.
Thanks for your help in advance!
|