Hi Salvatore,
> does anyone know the difference between how HRF deconvolution is implemented in the PPI analysis stream vs. the DCM stream? are they identical, is the latter an elaboration of the former, or are they different beasts entirely? thank you
They're different beasts entirely. The deconvolution in the PPI
analysis stream assumes a canonical HRF. By contrast, DCM contains
parameters for modeling the neurovascular coupling, and can
effectively come up with a different HRF for each of the regions
you're modeling based on fitting these parameters. This is
computationally more expensive, but presumably more accurate.
Hope this helps!
Best regards,
Jonathan
|