align_epi_anat.py question

[JW1]

Dear AFNI experts,

I ran align_epi_anat.py below which generated align_tlrc.jpg attached. For this, both anat (T1) and epi files were generated as tlrc format.

align_epi_anat.py
-epi2anat
-anat T1w_acpc_dc_restore.nii.gz
-epi ${subj}_REST1_LR+orig
-epi_base 0
-volreg off
-cost lpa+ZZ
-big_move
-overwrite

Therefore, when running run the proc.py , it showed the error below, which is the same as here (3dBucket error: output dataset name 'filename" conflicts with existing file)

3dbucket -prefix subj_T1.nii.gz
++ 3dbucket: AFNI version=AFNI_19.0.06 (Feb 4 2019) [64-bit]
** ERROR: output dataset name ‘subj_T1.nii.gz’ conflicts with existing file
** ERROR: dataset NOT written to disk!

Here is the proc py script I used.

afni_proc.py
-blocks despike align tlrc blur mask regress
-copy_anat /Volumes/Elements/RS/{subj}/sub_T1.nii.gz \ -anat_follower_ROI aaseg anat /Volumes/Elements/RS/{subj}/aparc.a2009s+aseg.nii.gz
-anat_follower_ROI aeseg epi /Volumes/Elements/RS/{subj}/aparc.a2009s+aseg.nii.gz \ -anat_follower_ROI FSvent epi /Volumes/Elements/RS/{subj}/vent.nii.gz
-anat_follower_ROI FSWm epi /Volumes/Elements/RS/{subj}/wm.nii.gz \ -anat_follower_erode FSvent FSWm \ -dsets /Volumes/Elements/RS/8/{subj}/{subj}_REST+orig \ -out_dir Proc_results_R2LR \ -script py_result_{subj}_script
-scr_overwrite
-subj_id ${subj}_V1_MR
-align_opts_aea -cost lpa -cost lpa+ZZ -giant_move
-tlrc_base MNI152_T1_2009c+tlrc
-tlrc_NL_warp
-tlrc_NL_awpy_rm no
-tlrc_no_ss
-regress_apply_mask
-regress_ROI_PC FSvent 3
-regress_make_corr_vols aeseg FSvent
-regress_anaticor_fast
-regress_anaticor_label FSWm
-regress_censor_motion 0.2
-regress_censor_outliers 0.1
-regress_apply_mot_types demean deriv
-regress_est_blur_epits
-regress_est_blur_errts

According to the suggestion from RIck in this thread (3dBucket error: output dataset name 'filename" conflicts with existing file), I changed the view to orig using this command line: setenv AFNI_NIFTI_VIEW orig

After that when running align_epi_anat.py using the same script, the alignment of the anat (T1) and epi files in orig are not as good as tlrc. (attached files including align orig_1). Could you please advise how to resolve this issue?

Best,
JW

[ptaylor]

Howdy-

I’m a little confused. Did you first run afni_proc.py, and now you are separately running align_epi_anat.py in the same directory? If so, I was wondering why you are running this separately, after the analysis is complete?

Separately, in this line of the afni_proc.py command:

-align_opts_aea -cost lpa -cost lpa+ZZ -giant_move \

… “-cost …” should not be specified twice. Also, generally neither lpa or lpa+ZZ should be used for EPI-anatomical alignment, because those two volumes have opposite tissue contrasts (NB: if a subject has an injection, sometimes that can change the relative EPI contrast, but such cases are rare). We would generally recommend using “lpc” or “lpc+ZZ” there, instead.

For your FreeSurfer follower ROIs, please check out this helpful page:
https://afni.nimh.nih.gov/pub/dist/doc/htmldoc/tutorials/fs/fs_makespec.html
In general, the REN datasets would be preferred to the “aparc.a2009s+aseg.nii.gz” ones. Also, there are WM and ventricle maps created for users in the output SUMA/ directory.

It is a little odd not to see a ‘volreg’ block, for motion estimation and reduction. I am not sure what this line:

-regress_apply_mot_types demean deriv \

… will do then, since there aren’t motion regressors to demean or derivatize?

-pt

[JW1]

Hi pt,

Thank you so much for the help! align epi anat was performed prior to the proc py. volreg was removed since using epi_al+tlrc generated an error message that volre should be removed in the proc py script.

I realized that SUMA on the freesurfer outputs wasn’t administered so performed SUMA and the alignment looks good. Thanks!

Best,
JW