Hello,

I have preprocessed fMRI data with CONN and would like to access the preprocessed voxel-level time series data.
However, the values in these files (range between -1 and 1) seem different from the values that usually consist BOLD time series (ranging between a few hundreds and a few thousands).
Would appreciate your advice on why that is the case (this is causing downstream issues such as very low SNR values).

Many thanks
Adi

Dear Alfonso,

I also have the same issue: I would like to use the confound-corrected time-series in a further analysis (outside of CONN), which expects the data in a positive range.

I tried switching the default BOLD signal units to "Raw" but the niftiData are still in the range {-500 500}. 


Is there an option to avoid centering around 0?

Alternatively, what would be the best way to re-center our data?

Would it suffice to estimate the static mean of the original data and then add that to the confound-corrected time-series, or would the mean have to be estimated and added to the images volume-per-volume?

Moreover, have the units been scaled in some way as a result of the (simultaneous) filtering/regression as well?

Many thanks,
Stavros