0ED6 reproduction

narps_open/core/interfaces/confounds.py

@@ -0,0 +1,100 @@
+#!/usr/bin/python
+# coding: utf-8
+
+""" This module defines custom interfaces related to confounds computation """
+
+from os.path import abspath
+from json import dump
+from itertools import zip_longest
+
+from nipype.interfaces.base import (
+    BaseInterface, BaseInterfaceInputSpec, traits, TraitedSpec, ImageFile, File
+    )
+from DVARS import DVARS_Calc
+
+class ComputeDVARSInputSpec(BaseInterfaceInputSpec):
+    """ Input specifications of a ComputeDVARS interface """
+    in_file = ImageFile(mandatory = True, desc = '4D nifti input file')
+    out_file_name = traits.Str(
+        mandatory = True,
+        desc = 'Base name for the output file, without extension'
+        )
+
+class ComputeDVARSOutputSpec(TraitedSpec):
+    """ Output specifications of a ComputeDVARS interface """
+    dvars_out_file = File(
+        exists = True,
+        desc = 'Output file containig DVARS results'
+        )
+    inference_out_file = File(
+        exists = True,
+        desc = 'Output file containig Inference results'
+        )
+    stats_out_file = File(
+        exists = True,
+        desc = 'Output file containig Stats results'
+        )
+
+class ComputeDVARS(BaseInterface):
+    """ Map the MATLAB code from the following article :
+
+    Afyouni, Soroosh & Nichols, Thomas. (2018).
+    Insight and inference for DVARS. NeuroImage. 172.
+    10.1016/j.neuroimage.2017.12.098.
+
+    Code is available here:
+    https://github.com/asoroosh/DVARS
+
+    Outputs a file containing DVARS computation at each time point of the in_file
+    """
+    input_spec = ComputeDVARSInputSpec
+    output_spec = ComputeDVARSOutputSpec
+
+    def _run_interface(self, runtime):
+        """ Run the DVARS computation and identify corrupted points """
+
+        # Compute DVARS
+        dvars = DVARS_Calc(self.inputs.in_file)
+
+        # Write results to DVARS file
+        with open(abspath(self.inputs.out_file_name + '_DVARS.tsv'), 'w') as file:
+            # Write header
+            file.write('DVARS\tDeltapDvar\tNDVARS_X2\n')
+
+            # Write data
+            for data in zip(
+                dvars['DVARS']['DVARS'].tolist(),
+                dvars['DVARS']['DeltapDvar'].tolist(),
+                dvars['DVARS']['NDVARS_X2'].tolist()):
+                file.write('\t'.join([str(e) for e in data]) + '\n')
+
+        # Write results to Inference file
+        with open(abspath(self.inputs.out_file_name + '_Inference.tsv'), 'w') as file:
+            # Write header
+            file.write('Pval\tH\tHStat\tHPrac\n')
+
+            # Write data
+            for data in zip_longest(
+                dvars['Inference']['Pval'].tolist(),
+                dvars['Inference']['H'].tolist(),
+                dvars['Inference']['HStat'][0].tolist(),
+                dvars['Inference']['HPrac'][0].tolist()):
+                file.write('\t'.join([str(e) for e in data]) + '\n')
+
+        # Write results to Stats file
+        with open(abspath(self.inputs.out_file_name + '_Stats.tsv'), 'w') as file:
+            # Write header
+            file.write('Mean\tSD\tDF\n')
+
+            # Write data
+            file.write(f"{dvars['Stats']['Mean']}\t{dvars['Stats']['SD']}\t{dvars['Stats']['DF']}\n")
+
+        return runtime
+
+    def _list_outputs(self):
+        outputs = self._outputs().get()
+        outputs['dvars_out_file'] = abspath(self.inputs.out_file_name + '_DVARS.tsv')
+        outputs['inference_out_file'] = abspath(self.inputs.out_file_name + '_Inference.tsv')
+        outputs['stats_out_file'] = abspath(self.inputs.out_file_name + '_Stats.tsv')
+
+        return outputs

narps_open/pipelines/__init__.py

@@ -10,7 +10,7 @@
 implemented_pipelines = {
     '08MQ': 'PipelineTeam08MQ',
     '0C7Q': None,
-    '0ED6': None,
+    '0ED6': 'PipelineTeam0ED6',
     '0H5E': None,
     '0I4U': None,
     '0JO0': None,