Hi,
I have resting state data which have been preprocesed (normalized, smoothed, regressed wm/csf/6motion parameters, temporal filtered).

1) I don't have their anatomical data, so I wonder structural inputs are necessary for an analysis.

2) If I want to calculate correlation between connectivity and behavior and remove age effect, is it ok to put
Allsubjects, behavior, Age [0 1 0] for between-subject contrast?


Thanks!


YM Huh

Hi YM Huh

Regarding (1), anatomical data is only used in CONN for two purposes: a) to generate gray/white/csf masks (e.g. for aCompCor); and b) for display purposes (for all subject-specific displays). In the absence of anatomical data you may still generate gray/white/csf masks by running the 'functional segmentation' module during your preprocessing steps, and for display purposes it is perfectly fine to use any MNI-space anatomical template. 

Regarding (2), yes, that is exactly correct.

Hope this helps
Alfonso
Originally posted by Youngmin Huh: