HVD NRG Nipype Interface to QA

nrg_interfaces.py

#!/usr/bin/env python
from nipype.interfaces.base import BaseInterface, BaseInterfaceInputSpec, TraitedSpec, traits, InputMultiPath, OutputMultiPath, File
from nipype.utils.filemanip import split_filename

import niftiqa
import os, sys

####### Begin Class niftiqa ########
class niftiqaInputSpec(BaseInterfaceInputSpec):
    niftis = InputMultiPath(traits.Either(traits.List(File(exists=True)),File(exists=True)),
                    desc='List of NIfTI scans to QA', mandatory=True, copyfile=False)

class niftiqaOutputSpec(TraitedSpec):
    reports = traits.List(OutputMultiPath(File(exists=True)),
                    desc="Lift of Image Reports, including mean, snr, and slice graphs.")

class niftiqaInterface(BaseInterface):
    input_spec = niftiqaInputSpec
    output_spec = niftiqaOutputSpec

    def _run_interface(self, runtime):
        args = niftiqa.parse_arguments(['--all'])
        for image in [os.path.basename(nifti) for nifti in self.inputs.niftis]:
            print image
            niftiqa.run(image, args)

        return runtime

    def _list_outputs(self):
        outputs = self._outputs().get()
        reports = []

        # Python has no native flatten(); here it's just a simple lambda.
        # This will only handle lists where everything is nested one level deep; 
        # see flatten_list in the main fmriNipype script for a more robust version using itertools
        flatten = lambda nested_list: [item for sublist in nested_list for item in sublist]
        
        for image in self.inputs.niftis:
            path, base, ext = split_filename(image)
            outfiles = [[
                os.path.join(path, base + '_history.html'),
                os.path.join(path, base + '_thumbnail_histogram.svg'),
                os.path.join(path, base + '_thumbnail_history.html'),
                os.path.join(path, base + '_thumbnail.png'),
                os.path.join(path, base + '_mean_slice.csv'),
                os.path.join(path, base + '_mean_slice.txt'),
                os.path.join(path, base + '_mean_slice.svg'),
                os.path.join(path, base + '_slice_report.txt'),
                os.path.join(path, base + '_slice_report.html')
            ]]

            for calculation in ['slope', 'snr', 'stdev', 'mask', 'mean']:
                outfiles.append([
                    os.path.join(path, base + '_%s_history.html'            % calculation ),
                    os.path.join(path, base + '_%s_thumbnail_histogram.svg' % calculation ),
                    os.path.join(path, base + '_%s_thumbnail_history.html'  % calculation ),
                    os.path.join(path, base + '_%s_thumbnail.png'           % calculation ),
                    os.path.join(path, base + '_%s.nii'                     % calculation )
                ])

            reports.append(flatten(outfiles))

                
        outputs['reports'] = reports
        return outputs