help > preprocessing physiological data
Showing 1-3 of 3 posts
Display:
Results per page:
Jun 6, 2013  11:06 PM | Vincent Beliveau
preprocessing physiological data
Good day,

For another study, I was previously given physiological data formatted to be compatible with CONN (one sample per frame of MRI data). I am processing a new study now and I only have access to the raw physiological data, sampled at a much higher rate than the MRI data. I was thinking of downsampling the data using PhLEM toolbox, however I was wondering if there was a better way to preprocess the physiological data to obtain optimal results with the CompCorr algorithm used by CONN.
Any suggestions?

Kind regards,
Vincent.
Jun 8, 2013  10:06 PM | Alfonso Nieto-Castanon - McGovern Institute for Brain Research. MIT
RE: preprocessing physiological data
Hi Vincent,

The CompCor algorithm (Behzadi et al. 2007) implemented in the CONN toolbox does not require obtaining any subject physiological data, as it derives potential physiological as well as movement effects on the BOLD timeseries by evaluating the signal within 'noise' ROIs (white matter and CSF areas). That said, if you have physiological data, I believe one useful way to incorporate that into CONN would be to use the RETROICOR method (Glover et al. 2000), which derives a set of regressors optimized to capture most of the variance induced by physiological effects on the BOLD timeseries, and add those regressors as additional confound effects into the CONN toolbox (I have not used this so I do not know the details but I believe you should be able to use the PhLEM toolbox to get the RETROICOR-based regressors as well).

Hope this helps
Alfonso
 
Originally posted by Vincent Beliveau:
Good day,

For another study, I was previously given physiological data formatted to be compatible with CONN (one sample per frame of MRI data). I am processing a new study now and I only have access to the raw physiological data, sampled at a much higher rate than the MRI data. I was thinking of downsampling the data using PhLEM toolbox, however I was wondering if there was a better way to preprocess the physiological data to obtain optimal results with the CompCorr algorithm used by CONN.
Any suggestions?

Kind regards,
Vincent.
Aug 10, 2018  01:08 AM | Amr Farahat - OVGU
RE: preprocessing physiological data
Hello Alfonso,

Is it okay to add the RETROICOR regressors after the preprocessing pipeline? or they have to be regressed before some preprocessing steps like slice time correction and spatial smoothing?
Does CONN perform the denoising step on the dataset 0 (smoothed) or dataset 1 (not smoothed) or both ? 
Thanks in advance.

Kind regards,
Amr