Hi Guys,
I used the DPARSF to process the data. Now, a reviewer commented on my manuscript as following:
"There has been no control for movement-related artifacts, which are now considered essential to control for in resting-state analysis. Movement parameters are not reported or compared across subjects. Censoring movement frames is strongly recommended at an FD 0.5 or lower level (Power, 2012; Van Dijk, 2011). "
As you know, i am not familiar with the data processing process. Thus, please let us know how to answer this question, or how to control the movement related artifacts.
Have a nice day.
Forums
Hi Henry,
Hi Henry,
You can cite Friston et al., 1996, Satterthwaite et al., Neuroimage 2013b and Yan et al., Neuroimage 2013 if you checked "Friston 24".
You can use mean FD as a covariate in group analysis, as suggested by Van Dijk et al., 2012, Satterthwaite et al., Neuroimage 2013a, Yan et al., Neuroimage 2013. For scrubbing, it is not so well accepted in the field, you can find argument against scrubbing in these papers. Although, if possible, you can put the scrubbing results into supplementary results -- DPARSF provides a convenient way to do scrubbing analysis. Basically you can check "scrubbing regressors".
Hope this helps.
Best,
Chao-Gan
Please like us at http://facebook.com/rfmri if you have a facebook account. Thanks! :)
Hi Chaogan,
Hi Chaogan,
Thanks a lot for the answer.
The 'scrubbing' you mentioned here is equal to 'movement-related artifactes'?
Hi Henry,
Hi Henry,
"Censoring movement frames" (which is what your reviewer suggested you to do to your data) is scrubbing, which means revomal of data points from timeseries if the data points were collected when subject was moving.
Your reviewer wanted you to address the motion effect on your data and suggested you to scrub your data. To do so, you can follow Chao-gan's suggestion and write about how you did motion correction at individule level, how you accounted for the motion effect at group level, and maybe show some evidence that your groups are not different in terms of motion. You can certainly scrub (or censor) your data and show how scrubbing affects your results as the reviwer suggested, but please keep in mind that scrubbing is a controveral data processing step which is on debating.
Hope this helps!
Yang