the same code run on different datasets -> +tlrc OR +orig produced

Dear AFNI experts,

I am using the same code for different datasets and for some datasets the results are in original space (files have the +orig extension), while for some datasets the results seem to be in Talairach space (there are the +tlrc extensions). I am using AFNI 16.2.02 and the code is:

     afni_proc.py                                                                \
        -subj_id $subject                                                        \
        -script proc.$subject -scr_overwrite                                     \
        -regress_opts_3dD -force_TR $TR                                          \
        -tcat_remove_first_trs 0                                                 \
        -dsets ${path_data}/${subject}_${task}_bold.nii                          \
        -volreg_align_to third                                                   \
        -regress_polort 2                                                        \
        -regress_bandpass 0.01 10                                                \
        -blur_size ${smoothing}.${zero_aux}                                      \
        -regress_stim_types AM1                                                  \
        -regress_stim_times $stim_times                                          \
        -regress_stim_labels activation_stimulus                                 \
        -regress_basis ${HRF}                                                    \
        -regress_make_ideal_sum sum_ideal.1D                                     \
        -regress_run_clustsim no                                                 \
        -regress_est_blur_epits                                                  \
        -regress_est_blur_errts                                                  \
        -regress_reml_exec                                                       \
        -regress_opts_reml -Rwherr whitened_errts.${subject}_REML

Do you have any idea why there might be differences?

Sorry for bothering you!

Best,

Wiktor Olszowy

I suspect that some of your input files are incorrectly labeled as being in standard space (+tlrc) in the NIFTI header. You can check this using 3dinfo.

I find it useful to explicitly define the blocks using a -blocks or -do_blocks command in afni_proc.py. Also if you want all of your data to end up in standard space, you likely want a -volreg_tlrc_warp in your command.

We usually suggest you start with one of the examples (like 6) for processing needs, these often have “best practices” built in. It’s worth noting that bandpass filtering isn’t usually used on task data.

-Peter

Dear Peter,

Thanks a lot for your response! Indeed, the problem was with the header (for the datasets with resulting +tlrc files, the headers also had wrong TR, which is why I used ‘-force_TR’). I did not think that AFNI checks the space code from header (using sform_code?) and then changes the output file names accordingly, but actually it makes sense. Regarding ‘-volreg_tlrc_warp’, at this stage I would like to analyse the data in subject’s space.

Thanks again!

Best,

Wiktor Olszowy

You can use 3drefit to change the header of the NIFTI files (or of AFNI files). I usually go for double when I’m converting odd headers:


3drefit -view orig -space ORIG myFile.nii.gz

Thanks a lot!