Skip to content

pet-linear - Linear processing of PET images

This pipeline performs spatial normalization to the MNI space and intensity normalization of PET images.

Its steps include:

  • clipping of the background to remove noisy values which can deteriorate further processing.
  • affine registration to the MNI152NLin2009cSym template [Fonov et al., 2011, 2009] in MNI space with the SyN algorithm [Avants et al., 2008] from the ANTs software package [Avants et al., 2014]
  • intensity normalization using the average PET uptake in reference regions resulting in a standardized uptake value ratio (SUVR) map
  • cropping of the registered images to remove the background

Clinica & BIDS specifications for PET modality

Since Clinica v0.6, PET data following the official specifications in BIDS version 1.6.0 are now compatible with Clinica. See BIDS page for more information.

Prerequisites

You need to have performed the t1-linear pipeline on your T1-weighted MR images.

Dependencies

If you only installed the core of Clinica, this pipeline needs the installation of ANTs on your computer.

Running the pipeline

The pipeline can be run with the following command line:

clinica run pet-linear [OPTIONS] BIDS_DIRECTORY CAPS_DIRECTORY ACQ_LABEL
                       {pons|cerebellumPons|pons2|cerebellumPons2}

where:

  • BIDS_DIRECTORY is the input folder containing the dataset in a BIDS hierarchy
  • CAPS_DIRECTORY is the output folder containing the results in a CAPS hierarchy
  • ACQ_LABEL is the label given to the PET acquisition, specifying the tracer used (trc-<acq_label>).
  • The reference region is used to perform intensity normalization (i.e. dividing each voxel of the image by the average uptake in this region) resulting in a standardized uptake value ratio (SUVR) map. It can be cerebellumPons or cerebellumPons2 (used for amyloid tracers) and pons or pons2 (used for FDG). See PET introduction for more details about masks versions.

with specific options :

  • -ui/--uncropped_image : By default, output images are cropped to a fixed matrix size of 169×208×179, 1 mm isotropic voxels. This allows reducing the computing power required when training deep learning models afterwards. Use this option if you do not want to get cropped images.
  • --random_seed : By default, results are not deterministic. Use this option if you want to obtain a deterministic output. The value you set corresponds to the random seed used by ANTs. This option requires ANTs version 2.3.0 onwards and is also compatible with ANTsPy.
  • -rec/--reconstruction_method: Select only images based on a specific reconstruction method.
  • --save_pet_in_t1w_space: Use this option to save the PET image in the T1w space after rigid transformation.
Optional parameters common to all pipelines
  • -tsv / --subjects_sessions_tsv

This flag allows you to specify in a TSV file the participants belonging to your subset. For instance, running the FreeSurfer pipeline on T1w MRI can be done using :

clinica run t1-freesurfer BIDS_PATH OUTPUT_PATH -tsv my_subjects.tsv

participant_id  session_id
sub-CLNC0001    ses-M000
sub-CLNC0001    ses-M018
sub-CLNC0002    ses-M000

Creating the TSV

To make the display clearer the rows here contain successive tabs but that should not happen in an actual TSV.

  • -wd / --working_directory

By default when running a pipeline, a temporary working directory is created. This directory stores all the intermediary inputs and outputs of the different steps of the pipeline. If everything goes well, the output directory is eventually created and the working directory is deleted.

With this option, a working directory of your choice can be specified. It is very useful for the debugging process or if your pipeline crashes. Then, you can relaunch it with the exact same parameters which will allow you to continue from the last successfully executed node. For the pipelines that generate many files, such as dwi-preprocessing (especially if you run it on multiple subjects), a specific drive/partition with enough space can be used to store the working directory.

  • -np / --n_procs

This flag allows you to exploit several cores of your machine to run pipelines in parallel, which is very useful when dealing with numerous subjects and multiple sessions. Thanks to Nipype, even for a single subject, a pipeline can be run in parallel by exploiting the cores available to process simultaneously independent sub-parts. We recommend using your_number_of_cpu - 1 for costly pipelines such as pet-surface-longitudinal.

If you do not specify -np / --n_procs flag, Clinica will detect the number of threads to run in parallel and propose the adequate number of threads to the user.

  • -cn / --caps-name

Use this option if you want to specify the name of the CAPS dataset that will be used inside the dataset_description.json file, at the root of the CAPS folder (see CAPS Specifications for more details). This works if this CAPS dataset does not exist yet, otherwise the existing name will be kept.

Several PET scans

It can happen that a BIDS dataset contains several PET scans for a given subject and session. In this situation, these images will differ through at least one BIDS entity like the tracer or the reconstruction method. When running the PET pipeline, clinica will raise an error if more than one image matches the criteria provided through the command line. To avoid that, it is important to specify values for these options such that a single image is selected per subject and session.

Tip

Do not hesitate to type clinica run pet-linear --help to see the full list of parameters.

Outputs

Results are stored in the following folder of the CAPS hierarchy: subjects/<participant_id>/<session_id>/pet_linear.

The main output files are:

  • <source_file>_space-MNI152NLin2009cSym_desc-Crop_res-1x1x1_suvr-<label>_pet.nii.gz: PET SUVR image registered to the MNI152NLin2009cSym template and cropped.
  • <source_file>_space-T1w_rigid.mat: rigid transformation between the PET and T1w images estimated with ANTs.
  • (optional) <source_file>_space-MNI152NLin2009cSym_res-1x1x1_pet.nii.gz: PET SUVR image affinely registered to the MNI152NLin2009cSym template (i.e. not cropped).
  • (optional) <source_file>_space-T1w_pet.nii.gz: PET image affinely registered to the associated T1w image.

Going further

You can now use the ClinicaDL framework presented in [Wen et al., 2020] for classification based on deep learning methods.

Describing this pipeline in your paper

Example of paragraph

These results have been obtained using the pet-linear pipeline of Clinica [Routier et al., 2021]. This pipeline first performs intra-subject rigid registration of the PET image into the space of the subject’s T1-weighted (T1w) MR image using the SyN algorithm [Avants et al., 2008] from ANTs [Avants et al., 2014]. The PET to T1w transformation is then composed with T1w to ICBM 2009c nonlinear symmetric template affine transformation, obtained with t1-linear, to transport the PET image to the MNI space [Fonov et al., 2011, 2009]. The PET image is further intensity normalized using the average PET uptake in a reference region ([pons | pons + cerebellum]), resulting in a standardized uptake value ratio (SUVR) map. The PET SUVR image in MNI space is finally cropped to remove the background, resulting in images of size 169×208×179 with 1 mm isotropic voxels.

Tip

Easily access the papers cited on this page on Zotero.

Contact us !