Hi David,
This is simply done by using the appropriate weights as parametric modulators. SPM will not offer you out of the box a selection of basis functions to use as parametric modulators.
Practically, this consists of assigning parametric weights for individual stimulus events according to a set of basis functions of your choice. In the paper, we used two functions to model the independent contributions of a Gaussian function, and another one that is sensitive to the width of the Gaussian. That is, for every single stimulus we had 2 parametric values to use as pmod. Depending on the basis function set you would like to use, this number can be different. However, there is essentially no difference between using a basis function set or any other type of parametric modulators to implement this in SPM.
To see how this is done in the batch mode, please refer to
https://github.com/selimonat/fancycarp/blob/mrt/fearamy/Subject.m#L2195
as you see the dummy.cond variable contains all the important information about the parametric predictors. They are computed in
https://github.com/selimonat/fancycarp/blob/mrt/fearamy/Subject.m#L2452
cheers,
S.
|