Group comparison, ROI analysis and smoothing

Hi @all,

I have a question regarding the smoothing of RS fMRI data. I want to compare patiens and controls based on ALFF, fALFF,  ReHO, VHMC and also want to use the extracted AAL ROIs to analyze each subject individually for calculating cross-correlation between ROIs. Now I was wondering if it is even necessary to apply smoothing to the data at all? I was also wondering if it is necessary to apply the head motion scrubbing regressors or if ths will corrupt the calculation of ALFF, fALL asf. ? Because as I understood scrubbing removes bad time points.

Can anyone help me out here?



Hi David,

For ReHo, smoothing in preprocessing should not be performed. For the others, you can smooth in preprocessing or after the derivatives were generated.

I don't recommend scrubbing for ALFF/fALFF analyses. For the others, supplementary analysis with scrubbing can be performed to confirm the main results.



Thank you :). Is the scrubbing option in DPARSF the same as including the head motion scrubbing regressors? In other words, should I use neither of those options for ALFF/fALFF or can I use the head motion scrubbing regressors -option instead of the scrubbing option to get better data?

------------------ 原始邮件 ------------------
发件人: "The R-fMRI Network"<>;
发送时间: 2015年2月22日(星期天) 凌晨0:09
收件人: ""<>;
主题: Re: [RFMRI] Group comparison, ROI analysis and smoothing

[To post a comment, please reply to

Hi David,

Scrubbing regressors are spike regression -> model the bad points in a regression model.

The second scrubbing is simply delete the bad points.

I will suggest the first option.



Thank you :)