Hey everybody.

Sorry for the basic question, somehow i dont't figure out the answer myself.
What does conn do with the identified outliers? In the default pipeline the threshold is 0.9. BUT, is their a threshold for removing the whole subject like when there are more than 50 outliers?? can i change that? i work with the batch and gui...

does conn displace the outliers with a mean of the previous and following timepoint? where can i read more about that??

thanks in advance!!

Till from Berlin

Hi Till

CONN's default denoising procedure does not interpolate outlier scans, nor does it remove entire subjects, it simply disregards individual scans/timepoints flagged as a potential outliers by virtue of explicitly regressing out any outlier effects as part the denoising procedure. For a general description of the denoising procedure in CONN see https://web.conn-toolbox.org/fmri-method... , and/or the "treatment of temporal confounding factors" section in the original CONN paper (listed in https://web.conn-toolbox.org/resources/documentation ), and for a general description/justification of this "scrubbing" procedure see Power et al. 2014 "Methods to detect, characterize, and remove motion artifact in resting state fMRI" manuscript)

Best
Alfonso
Originally posted by Till Langhammer:

For reference (spoilers: CONN has no issue): Eklund, A., Nichols, T., Afyouni, S., & Craddock, C. (2020). How does group differences in motion scrubbing affect false positives in functional connectivity studies?. BioRxiv. https://www.biorxiv.org/content/10.1101/2020.02.12.944454v1.abstract