error with 3dpc during afni_proc

AFNI Message Board
1

Hi,
I encountered the following error with several participants today when their proc scripts reached 3dpc. I haven’t seen it before so I don’t know what needs to be fixed. Any ideas?

Error: 3dpc -mask follow_ROI_FSvent+tlrc -pcsave 3 -prefix rm.ROIPC.FSvent rm.det_pcin_rall+tlrc
++ 3dpc: AFNI version=AFNI_17.2.09 (Aug 22 2017) [64-bit]
** ERROR: covariance diagonal (1,1) = 0
** ERROR: covariance diagonal (2,2) = 0
etc
** ERROR: covariance diagonal (441,441) = 0
** ERROR: covariance diagonal (442,442) = 0
** FATAL ERROR: *** program exiting right here and now!

Here’s my afni_proc.py script:

==========================================================================

script generated by the command:

afni_proc.py -subj_id C3537997 -script proc.C3537997 -scr_overwrite -blocks \

despike tshift align tlrc volreg blur mask regress -copy_anat \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/3537997_SurfVol.nii \

-anat_follower_ROI aaseg anat \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/aparc.a2009s+aseg.nii \

-anat_follower_ROI aeseg epi \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/aparc.a2009s+aseg.nii \

-anat_follower_ROI FSvent epi \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/FT_vent.nii \

-anat_follower_ROI FSWe epi \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/FT_WM.nii \

-anat_follower_erode FSvent FSWe -dsets \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/nv_3537997_20141015_resting_state+orig.HEAD \

/gpfs/gsfs6/users/gowinjl/new_analyses/3537997/nv_3537997_20141103_resting_state+orig.HEAD \

-tcat_remove_first_trs 2 -tlrc_base MNI_caez_N27+tlrc -tlrc_NL_warp \

-volreg_align_to MIN_OUTLIER -volreg_warp_dxyz 3.5 -volreg_align_e2a \

-volreg_tlrc_warp -regress_ROI_PC FSvent 3 -regress_make_corr_vols \

aeseg FSvent -regress_anaticor_fast -regress_anaticor_label FSWe \

-regress_censor_motion 0.2 -regress_censor_outliers 0.1 \

-regress_apply_mot_types demean deriv -regress_est_blur_epits \

-regress_est_blur_errts -regress_run_clustsim yes

2

That is strange. What is the full output of this?

3dmaskave -mask follow_ROI_FSvent+tlrc rm.det_pcin_rall+tlrc

You do not need to include the entire output time series.

Thanks,

  • rick
3

Hi Rick,
Here’s the output:

[gowinjl@biowulf C3537997.results]$ 3dmaskave -mask follow_ROI_FSvent+tlrc rm.det_pcin_rall+tlrc
++ 3dmaskave: AFNI version=AFNI_17.2.10 (Aug 30 2017) [64-bit]
+++ 306 voxels survive the mask
0 [306 voxels]
0 [306 voxels]
0 [306 voxels]
0 [306 voxels]
0 [306 voxels]
0 [306 voxels]
0 [306 voxels]
etc…

I don’t understand why those values would be 0

4

As a follow up, I looked at the following:

[gowinjl@biowulf C3537997.results]$ 3dmaskave -mask follow_ROI_FSWe+tlrc rm.det_pcin_rall+tlrc
++ 3dmaskave: AFNI version=AFNI_17.2.10 (Aug 30 2017) [64-bit]
+++ 9353 voxels survive the mask
0.237365 [9353 voxels]
0.070627 [9353 voxels]
-0.0387127 [9353 voxels]
0.025132 [9353 voxels]
0.125903 [9353 voxels]
etc

So the 0 values are only showing up in the FSvent mask, but I still don’t know why

5

The attached image shows the problem (the values are all at the top of the image) but I don’t know where this may have gone wrong.

rm.det_pcin_rall.jpg

6

That suggests registration between the anat and EPI failed. If you
view the original anat and EPI, how well do they overlay each other?
Look at the data input to afni_proc.py, for example.

  • rick
7

You’re right, that’s where the problem originated. I’ll see if I can resolve it. Thank you.