3 ms·
Without looking into this a lot, probably not much. This parcellation is described as multi-modal, so it's not using only structural or only diffusion, but I do
by irremediable 10y ago
Without looking into this a lot, probably not much. This parcellation is described as multi-modal, so it's not using only structural or only diffusion, but I doubt it's using fMRI much if at all. Even if so, I'm sure they validated the parcellation against pre-existing parcellations. I'll double-check this tomorrow when I can access the full paper more easily.
- brainsturgeon 10y agoThe parcellations in the paper were defined in part on task-dependent activity. I'm asking a simpler question: did the HCP paper use the false-positive prone analysis software?