fMRI preprocssing

Recently, with my knowledge updated, I have a few questions about the order of rsfMRI preprocessing. 

1. about the nuisance covariates regression, some of my friends told me that this step should be done after normalization. But, as far as I see, the nuisance covariates regression (head motion, CSF, WM...) was removed before normalization in DPABI. therefore, I was confused, whether both processings are reasonable or different intrinsic setups are in different processing toolboxes? Would you please help me solve it out?

e.g. as suggested by GRETNA, researchers canarbitrarily designate the order of preprocessing steps (except for Regress Out Covariates, which extract time series based on an image mask in MNI space and thus must be executed after Normalization)

2. about using FunImgARCWDFSB to calculate graph theory, can I directly use FunImgARCWDFSB to construct network matrix or I should modulate the preprocessing order (e.g. FunImgARWDCFSB) before calculating ?

Looking forward to your reply.

Thank you so much.

All best!


1. Both are OK. We prefer exclude the nuisance signals at earlier stage.

2. FunImgARCWDFSB is fine.

Thank you for you time!