In doing PPI analyses the BOLD signal of your seed is deconvolved to
approximate a neuronal response that would produce such a BOLD response.
This is preferential to using seed BOLD response for PPI analyses because of
the arguments laid out in Gitelman, et al. Modeling regional and
psychophysiologic interactions in fMRI: the importance of hemodynamic
deconvolution. NeuroImage (2003).
For those same reasons wouldn't it make sense to also deconvolve the BOLD
signal throughout the entire brain then examine psychophysio interactions?
This way you are getting closer to interactions at a neuronal level which
would be more appropriate because the functinal coupling PPI attempts to
elucidate occurs at a neuronal level, not a hemodynamic level, right?
I ask because my impression is SPM2 does not deconvolve anything more than
the seed and I'm not sure why. Is it an issue of memory since the
approximation is done per .1250 seconds and over the entire brain that would
create very large data sets or does it have to do with too many approximations?
I trust the current PPI setup is well-founded I guess I'm just having
trouble understanding this aspect. Thanks in advance.
Patrick
|