AFNI QC Rápido

Este tutorial es para realizar un control de calidad inicial de tus datos de rsfMRI o task fMRI usando AFNI.

Crear un folder individual para QC, yo cree uno llamado afniqc.

Dentro, puedes correr los comandos, o mi preferencia es crear un pequeño script, que llamé afniqc:

ap_run_simple_rest.tcsh -subjid sub-603.qc \
          -run_proc -nt_rm 2                \
          -anat $curso/data_00_basic/sub-603/ses-01/anat/sub-603_ses-01_run-01_T1w.nii.gz          \
          -epi $curso/data_00_basic/sub-603/ses-01/func/sub-603_ses-01_task-rest_run-01_bold.nii.gz

El comando ap_run_simple_rest.tcsh es muy sencillo pero puede modificarse:

------------------------------------------------------------------------------
ap_run_simple_rest.tcsh  - run a quick afni_proc.py analysis for QC

   usage: ap_run_simple_rest.tcsh [options] -anat ANAT -epi EPI1 EPI2 EPI3 ...

This program is meant to run a moderately quick single subject analysis,
treating the EPI as resting state data.

Overview:

   0. This program will change over time.  Do not rely on a fixed version.
      See "ap_run_simple_rest.tcsh -ver" for the current version number.
   1. Output files are placed in the current directory, so it is suggested
      to run from a "clean" location, generally away from the raw inputs.
   2. Template registration is merely affine, to be fast.
   3. By default, the first 2 time points are removed as pre-steady state.
      It is a good idea to set -nt_rm appropriately.

   inputs   : anat (optional), EPI

   controls : recommended opts: -subjid, -nt_rm

   outputs  : run_ap_SUBJID   - afni_proc.py command script
            : proc.SUBJID     -_proc script (if AP is run)
            : SUBJID.results  - proc results dir (if run)
            : out.*           - text output files from AP and proc scripts

This program may be devoured by afni_proc.py itself, at some point.

------------------------------------------------------------------------------
example 0: just create an afni_proc.py script, run_ap_SUBJ, no data required

      ap_run_simple_rest.tcsh -anat anat.nii -epi epi.nii


example 1: quickly process EPI (no anat, so no align/tlrc blocks)

      ap_run_simple_rest.tcsh -epi epi.nii -run_proc


example 2: preferred - run an analysis from a clean directory

      cd AFNI_data6/FT_analysis
      mkdir test.ap
      cd test.ap
      ap_run_simple_rest.tcsh -subjid ft.qc \
          -run_proc -nt_rm 2                \
          -anat ../FT/FT_anat+orig          \
          -epi ../FT/FT_epi_r*.HEAD

------------------------------------------------------------------------------
terminal options:

   -help                   : show this help
   -hist                   : show the program history
   -ver                    : show the version number

required parameters:

   -epi EPI_r1 EPI_r2 ...  : specify a list of EPI datasets

optional parameters:

   -anat ANAT              : specify single anatomical dataset

   -nt_rm NT               : num time points to remove from starts of runs
                             def: 2

   -run_ap                 : actually run the afni_proc.py command
                             def: do not, just generate AP command script

   -run_proc               : do the processing (run the proc script from AP)
                             def: do not run AP or proc script

   -subjid SUBJ_ID         : specify subject ID for file names
                             def: SUBJ

   -template TEMPLATE      : specify template for standard space
                             def: MNI152_2009_template_SSW.nii.gz

   -compressor COMP        : control automatic compression of *.BRIK files.
                             'COMP' must be one of the allowed keywords for
                             the AFNI_COMPRESSOR environment variable:
                                GZIP  COMPRESS  BZIP2  PIGZ
                             and you must have the associated program for
                             compression installed (e.g., 'gzip')
                             def: not set here

   -verb VERB              : specify verbosity level (3 == -echo)
                             def: 1

   -echo                   : set 'echo' in the shell, as if run via 'tcsh -x'
                             (same as '-verb 3')

------------------------------------------------------------------------------
R Reynolds Apr, 2021
version 0.6
------------------------------------------------------------------------------

Lo importante es saber que el template se puede modificar. Si se quieren hacer roedores u otros animales, es necesario agregar el -template TEMPLATE y quitar del script -run_proc para que no corra completo. Al correrlo va a crear el script final, y en ese hay que modificar el -blur a algo que sea para ratas por ejemplo (0.5 en lugar de 6 o 0 incluso).

Se corre el script afniqc, yo lo hago usando bash afniqc. En mi computadora y con estos datos tarda 1 hora y media por sujeto aproximadamente.

Al terminar nos da un folder, en este caso sub-603.qc.results. Dentro, hay que buscar el folder QC_sub-603.qc y el archivo index.html que tiene el resumen de información. Para abrirlo correctamente, se utiliza el comando:

open_apqc.py -infiles QC_sub-603.qc/index.html

Al abrirlo, podemos interactuar con la pagina y los cambios se guardan.