Suspicious results from SCA

Hello everyone,

we tried to run a seed-based resting-state functional connectivity analysis using feat. Even though we exactly followed the instructions provided on this website (FSL fMRI Resting State Seed-based Connectivity — neuroimaging core 0.1.1 documentation), our higher-level output indicates perfect correlations between the OFC (our seed region) and regions all across the brain, which seems very unlikely.

During the analysis, we started by extracting our ROI (i.e.: the OFC) from the harvardoxford atlas in 2mm thr50. After binarizing this mask, we put it into the functional space of each subject. Subsequently, the ROI timeseries was extracted using the following command:

Time series extraction


  •          -i "/Volumes/Frieda/rs-fc_analysis/Preprocessed/Resting_State/Addison_AllControls/$i/fmri_clean/denoised_func_data_nonaggr.nii.gz" \*
  •          -o "/Volumes/Frieda/rs-fc_analysis/Preprocessed/Resting_State/Addison_AllControls/$i/2mm_OFC.txt" \*
  •          -m "/Volumes/Frieda/rs-fc_analysis/Preprocessed/Resting_State/Addison_AllControls/$i/2mm_OFC_func_bin.nii.gz"* 

Then, we ran the first-level analysis and higher level, putting the wholebrain standard mask in ‘pre-threshold masking’ and received the suspicious results that you will find in the attachment. Since the first-level results already show correlations of the OFC with the whole brain we suspected that this part of the analysis might not correctly display the brain regions that show correlations with the mean OFC activity.

Since the OFCxwholebrain analysis appears suspicious we were wondering whether something within the feat analyses went wrong.

Any help on how to check whether the analysis went correctly/how to fix this issue would be very much appreciated!

Kindly, Frieda