Dear all,
We acquired two fMRI sessions (with two different stimuli) on each subject, and we would like to decide whether the "shape" of activation induced by the stimuli used in the first session is different from the "shape" of activation induced by the stimuli used in the second session. We would like to answer this question using ICA (because some paradigm correlated motion is assumed).
Unfortunately, the paradigms and the measurement-lengths were somewhat different between the two sessions, so multi-session tensor-ICA is not appropriate for comparison.
Maybe the multi-session temporal concatenation/dualreg is appropriate for the comparison, but we don't know whether multi-session temporal concatenation method and the dualreg script are able to handle sessions with different paradigms/lengths? Because the paradigms/lengths are somewhat different we are not interested in the "amplitude" of activation only the "shape", therefore dualreg script would be used without variance-normalise the timecourses.
Is it able/appropriate to use the multi-session temporal concatenation/dualreg method for this purpose?
I'd appreciate any answer. Thank your answers in advance.
Thanks,
Gabor
|