Harmonize PET tracers handling (ML / DL / Stats)

clinica/pipelines/machine_learning/region_based_io.py

@@ -4,60 +4,6 @@
 import numpy as np
 import pandas as pd
 import nibabel as nib
-from os.path import join
-
-
-def get_caps_t1_list(input_directory, subjects_visits_tsv, group_label, atlas_id):
-    """
-    path to arrive to the list of the file with the statistics on atlas_id
-    Args:
-        input_directory:
-        subjects_visits_tsv:
-        group_label:
-        atlas_id:
-
-    Returns:
-
-    """
-
-    from os.path import join
-    import pandas as pd
-
-    subjects_visits = pd.io.parsers.read_csv(subjects_visits_tsv, sep='\t')
-    if list(subjects_visits.columns.values) != ['participant_id', 'session_id']:
-        raise Exception('Subjects and visits file is not in the correct format.')
-    subjects = list(subjects_visits.participant_id)
-    sessions = list(subjects_visits.session_id)
-    image_list = [join(input_directory + '/subjects/' + subjects[i] + '/'
-                       + sessions[i] + '/t1/spm/dartel/group-' + group_label + '/atlas_statistics/' + subjects[i] + '_'
-                       + sessions[i]+'_T1w_space-'+atlas_id+'_map-graymatter_statistics.tsv')
-                  for i in range(len(subjects))]
-    return image_list
-
-
-def get_caps_pet_list(input_directory, subjects_visits_tsv, group_label, atlas_id):
-    """
-
-    Args:
-        input_directory:
-        subjects_visits_tsv:
-        group_label:
-        atlas_id:
-
-    Returns:
-
-    """
-
-    subjects_visits = pd.io.parsers.read_csv(subjects_visits_tsv, sep='\t')
-    if list(subjects_visits.columns.values) != ['participant_id', 'session_id']:
-        raise Exception('Subjects and visits file is not in the correct format.')
-    subjects = list(subjects_visits.participant_id)
-    sessions = list(subjects_visits.session_id)
-    image_list = [join(input_directory, 'analysis-series-default/subjects/' + subjects[i] + '/'
-                       + sessions[i] + '/pet/atlas_statistics/' + subjects[i] + '_' + sessions[i]
-                       + '_space-' + atlas_id + '_map-fdgstatistic2.tsv')
-                  for i in range(len(subjects))]
-    return image_list
 
 
 def load_data(image_list, subjects):
@@ -103,12 +49,14 @@ def features_weights(image_list, dual_coefficients, sv_indices, scaler=None):
     """
 
     if len(sv_indices) != len(dual_coefficients):
-        print("Length dual coefficients: " + str(len(dual_coefficients)))
-        print("Length indices: " + str(len(sv_indices)))
-        raise ValueError('The number of support vectors indices and the number of coefficients must be the same.')
+        raise ValueError(
+            f"The number of support vectors indices and the number of coefficients must be the same.\n"
+            f"- Number of dual coefficients: {len(dual_coefficients)}\n"
+            f"- Number of indices:: {len(sv_indices)}\n"
+        )
 
     if len(image_list) == 0:
-        raise ValueError('The number of images must be greater than 0.')
+        raise ValueError("The number of images must be greater than 0.")
 
     sv_images = [image_list[i] for i in sv_indices]
 
@@ -134,9 +82,6 @@ def weights_to_nifti(weights, atlas, output_filename):
     Returns:
 
     """
-
-    from os.path import join, split, realpath
-
     from clinica.utils.atlas import AtlasAbstract
 
     atlas_path = None

clinica/pipelines/machine_learning/voxel_based_io.py

@@ -1,44 +1,7 @@
 # coding: utf8
 
 import numpy as np
-import pandas as pd
 import nibabel as nib
-from os.path import join
-
-
-def get_caps_t1_list(input_directory, subjects_visits_tsv, group_label, fwhm, modulated):
-
-    subjects_visits = pd.io.parsers.read_csv(subjects_visits_tsv, sep='\t')
-    if list(subjects_visits.columns.values) != ['participant_id', 'session_id']:
-        raise Exception('Subjects and visits file is not in the correct format.')
-    subjects = list(subjects_visits.participant_id)
-    sessions = list(subjects_visits.session_id)
-    if fwhm == 0:
-        image_list = [join(input_directory, 'subjects/' + subjects[i] + '/'
-                           + sessions[i] + '/t1/spm/dartel/group-' + group_label + '/'
-                           + subjects[i] + '_' + sessions[i] + '_T1w_segm-graymatter'+'_space-Ixi549Space_modulated-'+modulated+'_probability.nii.gz') for i in range(len(subjects))]
-    else:
-        image_list = [join(input_directory, 'subjects/' + subjects[i] + '/'
-                           + sessions[i] + '/t1/spm/dartel/group-' + group_label + '/'
-                           + subjects[i] + '_' + sessions[i] + '_T1w_segm-graymatter' + '_space-Ixi549Space_modulated-' + modulated + '_fwhm-'+fwhm+'mm_probability.nii.gz')
-                      for i in range(len(subjects))]
-
-    return image_list
-
-
-def get_caps_pet_list(input_directory, subjects_visits_tsv, group_label, pet_type):
-
-    subjects_visits = pd.io.parsers.read_csv(subjects_visits_tsv, sep='\t')
-    if list(subjects_visits.columns.values) != ['participant_id', 'session_id']:
-        raise Exception('Subjects and visits file is not in the correct format.')
-    subjects = list(subjects_visits.participant_id)
-    sessions = list(subjects_visits.session_id)
-
-    image_list = [join(input_directory, 'subjects/' + subjects[i] + '/'
-                       + sessions[i] + '/pet/preprocessing/group-' + group_label + '/' + subjects[i]
-                       + '_' + sessions[i] + '_task-rest_acq-' + pet_type + '_pet_space-Ixi549Space_pet.nii.gz') for i in range(len(subjects))]
-
-    return image_list
 
 
 def load_data(image_list, mask=True):

clinica/pipelines/machine_learning_spatial_svm/spatial_svm_cli.py

@@ -18,6 +18,8 @@ def define_options(self):
         """Define the sub-command arguments."""
         from colorama import Fore
         from clinica.engine.cmdparser import PIPELINE_CATEGORIES
+        from clinica.utils.pet import LIST_SUVR_REFERENCE_REGIONS
+
         # Clinica compulsory arguments (e.g. BIDS, CAPS, group_label)
         clinica_comp = self._args.add_argument_group(PIPELINE_CATEGORIES['CLINICA_COMPULSORY'])
         clinica_comp.add_argument("caps_directory",
@@ -30,17 +32,24 @@ def define_options(self):
                                        'pet-volume' to use SUVr maps.''',
                                   choices=['t1-volume', 'pet-volume'],
                                   )
-        # Optional arguments
+        # Optional arguments for inputs from pet-volume pipeline
         optional_pet = self._args.add_argument_group(
-            '%sPipeline options if you use inputs from pet-volume pipeline%s' %
-            (Fore.BLUE, Fore.RESET)
+            f"{Fore.BLUE}Pipeline options if you use inputs from pet-volume pipeline{Fore.RESET}"
         )
-        optional_pet.add_argument("-pt", "--pet_tracer",
-                                  default='fdg',
-                                  help='PET tracer. Can be fdg or av45 (default: --pet_tracer %(default)s)')
-        optional_pet.add_argument("-no_pvc", "--no_pvc",
-                                  action='store_true', default=False,
-                                  help="Force the use of non PVC PET data (by default, PVC PET data are used)")
+        optional_pet.add_argument("-acq", "--acq_label",
+                                  type=str,
+                                  default=None,
+                                  help='Name of the label given to the acquisition, specifying the tracer used (acq-<acq_label>).')
+        optional_pet.add_argument("-suvr", "--suvr_reference_region",
+                                  choices=LIST_SUVR_REFERENCE_REGIONS,
+                                  default=None,
+                                  help='Intensity normalization using the average PET uptake in reference regions '
+                                       'resulting in a standardized uptake value ratio (SUVR) map. It can be '
+                                       'cerebellumPons (used for amyloid tracers) or pons (used for 18F-FDG tracers).')
+        optional_pet.add_argument("-pvc", "--use_pvc_data",
+                                  action='store_true',
+                                  default=False,
+                                  help="Use PET data with partial value correction (by default, PET data with no PVC are used)")
         # Clinica standard arguments (e.g. --n_procs)
         self.add_clinica_standard_arguments()
         # Advanced arguments (i.e. tricky parameters)
@@ -54,15 +63,33 @@ def define_options(self):
     def run_command(self, args):
         """Run the pipeline with defined args."""
         from networkx import Graph
+        from colorama import Fore
         from .spatial_svm_pipeline import SpatialSVM
+        from clinica.utils.exceptions import ClinicaException
         from clinica.utils.ux import print_end_pipeline, print_crash_files_and_exit
 
+        if args.orig_input_data == 'pet-volume':
+            if args.acq_label is None:
+                raise ClinicaException(
+                    f"{Fore.RED}You selected pet-volume pipeline without setting --acq_label flag. "
+                    f"Clinica will now exit.{Fore.RESET}"
+                )
+            if args.suvr_reference_region is None:
+                raise ClinicaException(
+                    f"{Fore.RED}You selected pet-volume pipeline without setting --suvr_reference_region flag. "
+                    f"Clinica will now exit.{Fore.RESET}"
+                )
+
         parameters = {
+            # Clinica compulsory arguments
             'group_label': args.group_label,
             'orig_input_data': args.orig_input_data,
-            'pet_tracer': args.pet_tracer,
-            'no_pvc': args.no_pvc,
-            'fwhm': args.fwhm,
+            # Optional arguments for inputs from pet-volume pipeline
+            'acq_label': args.acq_label,
+            'use_pvc_data': args.use_pvc_data,
+            'suvr_reference_region': args.suvr_reference_region,
+            # Advanced arguments
+            'fwhm': args.full_width_half_maximum,
         }
         pipeline = SpatialSVM(
             caps_directory=self.absolute_path(args.caps_directory),

clinica/pipelines/machine_learning_spatial_svm/spatial_svm_pipeline.py

@@ -16,18 +16,20 @@ def check_pipeline_parameters(self):
         """Check pipeline parameters."""
         from clinica.utils.group import check_group_label
 
-        if 'group_label' not in self.parameters.keys():
-            raise KeyError('Missing compulsory group_label key in pipeline parameter.')
+        # Clinica compulsory parameters
+        self.parameters.setdefault('group_label', None)
+        check_group_label(self.parameters['group_label'])
+
         if 'orig_input_data' not in self.parameters.keys():
             raise KeyError('Missing compulsory orig_input_data key in pipeline parameter.')
-        if 'fwhm' not in self.parameters.keys():
-            self.parameters['fwhm'] = 4
-        if 'pet_tracer' not in self.parameters.keys():
-            self.parameters['pet_tracer'] = 'fdg'
-        if 'no_pvc' not in self.parameters.keys():
-            self.parameters['no_pvc'] = False
 
-        check_group_label(self.parameters['group_label'])
+        # Optional parameters for inputs from pet-volume pipeline
+        self.parameters.setdefault('acq_label', None)
+        self.parameters.setdefault('suvr_reference_region', None)
+        self.parameters.setdefault('use_pvc_data', False)
+
+        # Advanced parameters
+        self.parameters.setdefault('fwhm', 4)
 
     def check_custom_dependencies(self):
         """Check dependencies that can not be listed in the `info.json` file.
@@ -53,14 +55,14 @@ def get_output_fields(self):
         return ['regularized_image']
 
     def build_input_node(self):
-        """Build and connect an input node to the pipeline.
-        """
+        """Build and connect an input node to the pipeline."""
         import os
         from colorama import Fore
         import nipype.pipeline.engine as npe
         import nipype.interfaces.utility as nutil
         from clinica.utils.inputs import clinica_file_reader, clinica_group_reader
-        from clinica.utils.input_files import t1_volume_final_group_template
+        from clinica.utils.input_files import (t1_volume_final_group_template,
+                                               pet_volume_normalized_suvr_pet)
         from clinica.utils.exceptions import ClinicaCAPSError, ClinicaException
         from clinica.utils.ux import print_groups_in_caps_directory
 
@@ -84,23 +86,26 @@ def build_input_node(self):
                 'description': 'graymatter tissue segmented in T1w MRI in Ixi549 space',
                 'needed_pipeline': 't1-volume-tissue-segmentation'
             }
-        elif self.parameters['orig_input_data'] is 'pet-volume':
-            if self.parameters['no_pvc']:
-                caps_files_information = {
-                    'pattern': os.path.join('pet', 'preprocessing', 'group-' + self.parameters['group_label'],
-                                            '*_pet_space-Ixi549Space_suvr-pons_pet.nii.gz'),
-                    'description': self.parameters['pet_tracer'] + ' PET in Ixi549 space',
-                    'needed_pipeline': 'pet-volume'
-                }
-            else:
-                caps_files_information = {
-                    'pattern': os.path.join('pet', 'preprocessing', 'group-' + self.parameters['group_label'],
-                                            '*_pet_space-Ixi549Space_pvc-rbv_suvr-pons_pet.nii.gz'),
-                    'description': self.parameters['pet_tracer'] + ' PET partial volume corrected (RBV) in Ixi549 space',
-                    'needed_pipeline': 'pet-volume with PVC'
-                }
+        elif self.parameters['orig_input_data'] == 'pet-volume':
+            if not (
+                self.parameters["acq_label"]
+                and self.parameters["suvr_reference_region"]
+            ):
+                raise ValueError(
+                    f"Missing value(s) in parameters from pet-volume pipeline. Given values:\n"
+                    f"- acq_label: {self.parameters['acq_label']}\n"
+                    f"- suvr_reference_region: {self.parameters['suvr_reference_region']}\n"
+                    f"- use_pvc_data: {self.parameters['use_pvc_data']}\n"
+                )
+            caps_files_information = pet_volume_normalized_suvr_pet(
+                acq_label=self.parameters["acq_label"],
+                suvr_reference_region=self.parameters["suvr_reference_region"],
+                use_brainmasked_image=False,
+                use_pvc_data=self.parameters["use_pvc_data"],
+                fwhm=0
+            )
         else:
-            raise ValueError('Image type ' + self.parameters['orig_input_data'] + ' unknown.')
+            raise ValueError(f"Image type {self.parameters['orig_input_data']} unknown.")
 
         try:
             input_image = clinica_file_reader(self.subjects,
@@ -133,14 +138,11 @@ def build_input_node(self):
         ])
 
     def build_output_node(self):
-        """Build and connect an output node to the pipeline.
-        """
+        """Build and connect an output node to the pipeline."""
         pass
 
     def build_core_nodes(self):
-        """Build and connect the core nodes of the pipeline.
-        """
-
+        """Build and connect the core nodes of the pipeline."""
         import clinica.pipelines.machine_learning_spatial_svm.spatial_svm_utils as utils
         import nipype.interfaces.utility as nutil
         import nipype.pipeline.engine as npe

clinica/pipelines/pet_surface/pet_surface_cli.py

@@ -17,19 +17,19 @@ def define_description(self):
     def define_options(self):
         """Define the sub-command arguments."""
         from clinica.engine.cmdparser import PIPELINE_CATEGORIES
+        from clinica.utils.pet import LIST_SUVR_REFERENCE_REGIONS
         # Clinica compulsory arguments (e.g. BIDS, CAPS, group_label)
         clinica_comp = self._args.add_argument_group(PIPELINE_CATEGORIES['CLINICA_COMPULSORY'])
         clinica_comp.add_argument("bids_directory",
                                   help='Path to the BIDS directory.')
         clinica_comp.add_argument("caps_directory",
                                   help='Path to the CAPS directory. (Filled with results from t1-freesurfer pipeline')
         clinica_comp.add_argument("acq_label", type=str,
-                                  help='Name of the PET tracer label in the acquisition entity '
-                                       '(acq-<acq_label>).')
-        clinica_comp.add_argument("suvr_reference_region",  choices=['cerebellumPons', 'pons'],
+                                  help='Name of the label given to the acquisition, specifying the tracer used (acq-<acq_label>).')
+        clinica_comp.add_argument("suvr_reference_region",  choices=LIST_SUVR_REFERENCE_REGIONS,
                                   help='Intensity normalization using the average PET uptake in reference regions '
                                        'resulting in a standardized uptake value ratio (SUVR) map. It can be '
-                                       'cerebellumPons (used for AV45 tracers) or pons (used for 18F-FDG tracers).')
+                                       'cerebellumPons (used for amyloid tracers) or pons (used for 18F-FDG tracers).')
         clinica_comp.add_argument("pvc_psf_tsv",
                                   help='TSV file containing for each PET image its point spread function (PSF) measured '
                                        'in mm at x, y & z coordinates. Columns must contain: '

clinica/pipelines/pet_surface/pet_surface_longitudinal_cli.py

@@ -29,7 +29,7 @@ def define_options(self):
         clinica_comp.add_argument("suvr_reference_region",  choices=['cerebellumPons', 'pons'],
                                   help='Intensity normalization using the average PET uptake in reference regions '
                                        'resulting in a standardized uptake value ratio (SUVR) map. It can be '
-                                       'cerebellumPons (used for AV45 tracers) or pons (used for 18F-FDG tracers).')
+                                       'cerebellumPons (used for amyloid tracers) or pons (used for 18F-FDG tracers).')
         clinica_comp.add_argument("pvc_psf_tsv",
                                   help='TSV file containing for each PET image its point spread function (PSF) measured '
                                        'in mm at x, y & z coordinates. Columns must contain: '

clinica/pipelines/pet_surface/pet_surface_pipeline.py

@@ -157,10 +157,10 @@ def build_input_node_longitudinal(self):
 
         check_relative_volume_location_in_world_coordinate_system('T1w-MRI (orig_nu.mgz)',
                                                                   read_parameters_node.inputs.orig_nu,
-                                                                  self.parameters['acq_label'].upper() + ' PET',
+                                                                  self.parameters['acq_label'] + ' PET',
                                                                   read_parameters_node.inputs.pet,
                                                                   self.bids_directory,
-                                                                  self.parameters['acq_label'].lower())
+                                                                  self.parameters['acq_label'])
 
         self.connect([
             (read_parameters_node, self.input_node, [('pet', 'pet'),
@@ -260,9 +260,9 @@ def build_input_node_cross_sectional(self):
             raise ClinicaException(error_message)
 
         check_relative_volume_location_in_world_coordinate_system('T1w-MRI (orig_nu.mgz)', read_parameters_node.inputs.orig_nu,
-                                                                  self.parameters['acq_label'].upper() + ' PET', read_parameters_node.inputs.pet,
+                                                                  self.parameters['acq_label'] + ' PET', read_parameters_node.inputs.pet,
                                                                   self.bids_directory,
-                                                                  self.parameters['acq_label'].lower())
+                                                                  self.parameters['acq_label'])
 
         self.connect([
             (read_parameters_node, self.input_node, [('pet',                 'pet'),