Using freesurfer ROIs from dtseries issue

Description:

I am trying to extract from dtseries the info from the freesurfer parcellation (e.g. hippocampus ROIs e.g. 17 and 53). From the documentation we should be able to specific the roi file (e.g. images/segmentation/freesurfer/mri/aparc+aseg_bold.nii.gz or aparc+aseg_t1.nii.gz) that we want to use in the input list file. However, I am getting 2 errors with either input file.

Call:

qunex_container fc_extract_roi_timeseries \
      --container='/gpfs/gibbs/pi/n3/software/Singularity/qunex_suite-1.0.3.sif' \
      --flist='/gpfs/gibbs/pi/n3/Studies/bihyper/processing/scripts/lists/gbc_pb15172_test.list' \
      --roiinfo='/gpfs/gibbs/pi/n3/Studies/bihyper/analysis/seeds/bihyper_hippocampus_CAB-NP_freesurfer.names' \
      --targetf="/gpfs/gibbs/pi/n3/Studies/bihyper/analysis/seeds/extract_hippocampus/" \
      --options="roimethod:mean|savegroup:long|saveind:none|eventdata=mean"

Logs:
No logs created.
Output errors:

  1. when using aparc+aseg_t1.nii.gz
    → command worked but outputs for left and right hippocampus are exactly the same.
  • Location of output: /gpfs/gibbs/pi/n3/Studies/bihyper/analysis/seeds/extract_hippocampus/extracted_hippocampus_CAB-NP_freesurfer.txt
  • Exemple for 1 session:
roi	V1_f1_max
CAB-NP-LeftHippocampus	0.019139
CAB-NP-RightHippocampus	0.016607
FS-LHipp	0.024058
FS-RHipp	0.024058
  1. when using aparc+aseg_bold.nii.gz
Processing session pb13736     ... reading image file(s) ... 1 frames read, done.
     ... creating ROI mask

Matlab Error! Processing Failed!
reshape: SIZE is not divisible by the product of known dimensions (= 91282)

Goal is to run this extraction and also run fc_compute_seedmap with this names file.

Thanks for your help!
Clara

Hi, I cannot test the code right now, but I think the problem might be that your roi file is in volume format (.nii.gz) and your file is in surface format (.dtseries.nii).

Do you have a solution to be able to use the freesurfer parcellation (for subcortical areas) to both extract and run seedmap on surface data? (using the individual sessions segmentation files).

Grega will look into it.

1 Like