Hi, I'm trying to use CONN to analyze resting-state connectivity in a dataset that is already fully preprocessed.
I followed the recommendations in another help thread to disable the denoising steps in my batch script:
%no denoising procedures
batch.Denoising.filter=[0 inf];
batch.Denoising.detrending=0;
batch.Denoising.despiking=0;
batch.Denoising.confounds={};
batch.Denoising.done=1;
batch.Densoiing.overwrite = 'Yes';
but when I ran the batch script, the output during the run read:
Step 1/7: Expanding conditions
100.0% (Condition rest)
Step 2/7: Importing conditions/covariates
100.0% (Subject 2 Session 4)
Step 3/7: Updating Denoising variables
100.0%
Step 4/7: Denoising functional data
Regression: White Matter (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: CSF (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: Effect of rest (2P) (1 dimensions; no polynomial expansion; added first-order derivative; non-filtered)
Order: Regression followed by filtering (RegBP)
Detrending: 0
Despiking: 0
Bandpass: [0 Inf]
100.0% (Subject 2) processed 1 files
Step 5/7: Denoising ROI data
Regression: White Matter (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: CSF (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: Effect of rest (2P) (1 dimensions; no polynomial expansion; added first-order derivative; non-filtered)
Order: Regression followed by filtering (RegBP)
Detrending: 0
Despiking: 0
Bandpass: [0 Inf]
100.0% (Subject 2) processed 1 files
Step 7/7: Updating Analysis variables
100.0%
This seems like CONN is still running the CompCor procedure and regressing out the effect of rest plus WM and CSF? This is confusing to me as I specifically told CONN to not use any confounds in the batch script. Any idea whether CONN is actually still running confound regression, and if so, how to block this behavior?
Thanks,
Eric
I followed the recommendations in another help thread to disable the denoising steps in my batch script:
%no denoising procedures
batch.Denoising.filter=[0 inf];
batch.Denoising.detrending=0;
batch.Denoising.despiking=0;
batch.Denoising.confounds={};
batch.Denoising.done=1;
batch.Densoiing.overwrite = 'Yes';
but when I ran the batch script, the output during the run read:
Step 1/7: Expanding conditions
100.0% (Condition rest)
Step 2/7: Importing conditions/covariates
100.0% (Subject 2 Session 4)
Step 3/7: Updating Denoising variables
100.0%
Step 4/7: Denoising functional data
Regression: White Matter (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: CSF (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: Effect of rest (2P) (1 dimensions; no polynomial expansion; added first-order derivative; non-filtered)
Order: Regression followed by filtering (RegBP)
Detrending: 0
Despiking: 0
Bandpass: [0 Inf]
100.0% (Subject 2) processed 1 files
Step 5/7: Denoising ROI data
Regression: White Matter (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: CSF (5P) (5 dimensions; no polynomial expansion; no temporal expansion; non-filtered)
Regression: Effect of rest (2P) (1 dimensions; no polynomial expansion; added first-order derivative; non-filtered)
Order: Regression followed by filtering (RegBP)
Detrending: 0
Despiking: 0
Bandpass: [0 Inf]
100.0% (Subject 2) processed 1 files
Step 7/7: Updating Analysis variables
100.0%
This seems like CONN is still running the CompCor procedure and regressing out the effect of rest plus WM and CSF? This is confusing to me as I specifically told CONN to not use any confounds in the batch script. Any idea whether CONN is actually still running confound regression, and if so, how to block this behavior?
Thanks,
Eric