Dear SPM-users
We are looking at data from an experiment with transient happenings that are
modeled with events. When selecting hrf-convolution, there are also options
to add a time-derivative of the hrf. What are the reasons for this (any
references)? Wouldn’t it be more reasonable to use a time-derivative of the
hrf together with a time-derivative of the BOLD-signal in order to remove
sustained effects?
Thank you in advance
|