sounds like a good idea - have you had any problems with distortions though? I noticed that the occipital lobes of the fmri generally do not match that of the highres very well. I guess to make it worthwile in terms of applicability to all subjects I'd need to dilate quite generously, not sure if I'd gain much. But maybe eroding the BET fmri mask would be another option.
In any case, what do the FSL people do? I looked at the RSNs from Smith 2009 et al that are available to download - they pretty closely match the MNI template, except for the occipital lobes.
If it were only for esthetics... but maybe someone could comment on the effects on the stats, ie does using a mask that extends the brain will have much influence on the IC forming process, and on the final stats.
Thanks very much for all comments!
Torsten
|