Hi, 

I am curious what the default CONN preprocessing and denoising pipelines do with the detected outliers? Do they remove them from the data entirely or are they made into statistical covariates? 

Thank you! 
Leyla Brucar

the latter (as part of linear regression of confounding effects during denoising), see https://web.conn-toolbox.org/fmri-methods/denoising-pipeline for details
Alfonso
Originally posted by leylabrucar:

Hi Alfonso,

Thank you for your swift response! 

I am currently running the default denoising pipeline and have run into an error on step 5/7:
"warning: no valid data for ROI atlas.Cereb7 r (Cerebelum 7b Right) in /Volumes/Leyla Drive/PILOTK2016/1701/Functional/waurun-4.nii (scan # 1/300)"

Any advice on how to rectify this issue? 
All the best,
Leyla

Hi Leyla,

Cereb7 is in the inferior portion of the cerebellum, and the error indicates that there is no functional data in this area (all NaN's or 0's). This might indicate that perhaps your functional data field of view might crop some portion of the cerebellum, or alternatively some problem with the coregistration and normalization of your functional data?

Best
Alfonso
Originally posted by leylabrucar:

Hi Alfonso, 

Thank you again for the quick response. 

I went back and checked the QA registration: functional data and outline of ROI atlas plots and have attached one of the subjects plots below. Would this be an indication of this warning I have received during the denoising step? 

I am new to performing MRI analyses and am wondering if you have any suggestions as to how to figure out where this error may originate from and how to best proceed.

Thank you!
Best,
Leyla