Clarification on Implementation of aCompCor in fMRIPrep

Hi all,

I have a brief question on aCompCor. When calculating, are the six motion parameters as well as their temporal derivatives and quadratic terms all used? Are global signals used?


Hi @John_Massa,

Here is the citation text from the boiler plate:

Additionally, a set of physiological regressors were extracted to allow for component-based noise correction (CompCor, Behzadi et al. 2007). Principal components are estimated after high-pass filtering the preprocessed BOLD time-series (using a discrete cosine filter with 128s cut-off) for the two CompCor variants: temporal (tCompCor) and anatomical (aCompCor). tCompCor components are then calculated from the top 2% variable voxels within the brain mask. For aCompCor, three probabilistic masks (CSF, WM and combined CSF+WM) are generated in anatomical space. The implementation differs from that of Behzadi et al. in that instead of eroding the masks by 2 pixels on BOLD space, a mask of pixels that likely contain a volume fraction of GM is subtracted from the aCompCor masks. This mask is obtained by dilating a GM mask extracted from the FreeSurfer’s aseg segmentation, and it ensures components are not extracted from voxels containing a minimal fraction of GM. Finally, these masks are resampled into BOLD space and binarized by thresholding at 0.99 (as in the original implementation). Components are also calculated separately within the WM and CSF masks. For each CompCor decomposition, the k components with the largest singular values are retained, such that the retained components’ time series are sufficient to explain 50 percent of variance across the nuisance mask (CSF, WM, combined, or temporal). The remaining components are dropped from consideration.

In short, none of the other regressors are applied, just a high-pass filter.


Thanks! I just have one follow up question then. This is a comment from the source code:

The confound time series derived from head motion estimates and global
signals were expanded with the inclusion of temporal derivatives and
quadratic terms for each [@confounds_satterthwaite_2013].

What is the “confound time series” referring to? I was hoping it was aCompCor as we are looking to have the PCs orthogonalized to the set of 24 motion parameters. I assume this is not an option in fMRIPrep.

It is just referring to the fact that they are time series that could be used in nuisance regression. Each head motion parameters (e.g., rotation in x direction) is a time series that could be a confound depending on your denoising strategy.

Got it. Thanks again!