Skip to content

Nipype workflow to generate fieldmaps from EPI acquisitions with differing phase-encoding directions

License

Notifications You must be signed in to change notification settings

andrewrosss/nipype-generate-fieldmaps

Repository files navigation

nipype-generate-fieldmaps

Nipype workflow to generate fieldmaps from EPI acquisitions with differing phase-encoding directions

Installation

pip install nipype-generate-fieldmaps

Usage

As a stand-alone workflow

# create the workflow
wf = create_generate_fieldmaps_wf()

# wire-up the inputs
wf.inputs.inputnode.se_epi_pe1_file = my_se_epi_pe1_file  # type: str | Path
wf.inputs.inputnode.se_epi_pe2_file = my_se_epi_pe2_file  # type: str | Path
wf.inputs.inputnode.se_epi_pe1_sidecar_file = my_se_epi_pe1_sidecar_file  # type: str | Path
wf.inputs.inputnode.se_epi_pe2_sidecar_file = my_se_epi_pe2_sidecar_file  # type: str | Path

# set the output directory
wf.base_dir = my_output_dir  # type: str | Path

# run it
wf.run()

As a nested workflow

The nodes node1, node2, some_other_node, maybe_a_4th_node, epi_node, and anat_node are made up for demonstration purposes

from nipype import Workflow
from nipype.interfaces.fsl import EpiReg
from nipype_generate_fieldmaps import create_generate_fieldmaps_wf

# parent workflow defined elsewhere
wf = Workflow(...)

# create the (sub-)workflow
fmap_wf = create_generate_fieldmaps_wf()

# connect the various nodes form the parent workflow to the nested fieldmap workflow
wf.connect(node1, 'out_file', fmap_wf, 'inputnode.se_epi_pe1_file')
wf.connect(node2, 'out', fmap_wf, 'inputnode.se_epi_pe2_file')
wf.connect(some_other_node, 'output_file', fmap_wf, 'inputnode.se_epi_pe1_sidecar_file')
wf.connect(maybe_a_4th_node, 'sidecar_file', fmap_wf, 'inputnode.se_epi_pe2_sidecar_file')

# connect the fieldmap workflow outputs to one (or more) node(s) in the parent workflow
# for example: EpiReg()
epireg = Node(EpiReg(out_base='epi2str.nii.gz'), name='epi_reg')
# from elsewhere
wf.connect(epi_node, 'my_epi_file' epireg, 'epi')
wf.connect(anat_node, 'my_t1_file', epireg, 't1_head')
wf.connect(anat_node, 'my_t1_brain_file', epireg, 't1_brain')
# from the fieldmap workflow!
wf.connect(fmap_wf, 'outputnode.fmap_rads_file', epireg, 'fmap')
wf.connect(fmap_wf, 'outputnode.fmap_mag_file', epireg, 'fmapmag')
wf.connect(fmap_wf, 'outputnode.fmap_mag_brain_file', epireg, 'fmapmagbrain')
wf.connect(fmap_wf, 'outputnode.echo_spacing', epireg, 'echospacing')
wf.connect(fmap_wf, 'outputnode.pe1_pedir', epireg, 'pedir')

From the command line

$ nipype-generate-fieldmaps --help
usage: nipype-generate-fieldmaps [-h] [-v] se_epi_pe1 se_epi_pe2 se_epi_pe1_sidecar se_epi_pe2_sidecar out_dir

Generate fieldmaps from EPI acquisitions with differing phase-encoding directions

positional arguments:
  se_epi_pe1          The spin-echo EPI file acquired in the 'first' phase-encoding direction
  se_epi_pe2          The spin-echo EPI file acquired in the 'second' phase-encoding direction
  se_epi_pe1_sidecar  The JSON sidecar for the first spin-echo EPI file
  se_epi_pe2_sidecar  The JSON sidecar for the second spin-echo EPI file
  out_dir             The directory into which outputs are written

optional arguments:
  -h, --help          show this help message and exit
  -v, --version       show program's version number and exit

Prerequisites

This workflow has a few requirements:

  1. There are two acquisitions (i.e. .nii.gz files) acquired with different phase encodings, usually opposite phase encodings but this need not be the case.

  2. The number of volumes in acquisition 1 (the first phase encoding direction) equals the number of volumes in acquisition 2 (the second phase encoding direction)

  3. Each acquisition has a JSON sidecar. Specifically, this workflow requires that each sidecar contain one of the following sets of properties. These properties are listed in the order in which the workflow will search:

    • PhaseEncodingDirection and TotalReadoutTime, or
    • PhaseEncodingDirection, ReconMatrixPE, and EffectiveEchoSpacing, or
    • PhaseEncodingDirection, ReconMatrixPE, and BandwidthPerPixelPhaseEncode

    If either JSON sidecar fails to contain at least one of the above sets of parameters the workflow will produce an error.

I/O

This workflow requires 4 inputs to be connected to the node named inputnode:

  • se_epi_pe1_file

    The spin-echo EPI file acquired in the 'first' phase-encoding direction

  • se_epi_pe2_file

    The spin-echo EPI file acquired in the 'second' phase-encoding direction

  • se_epi_pe1_sidecar_file

    The JSON sidecar for the first spin-echo EPI file

  • se_epi_pe2_sidecar_file

    The JSON sidecar for the second spin-echo EPI file

This workflow also exposes the following outputs via the node named outputnode:

  • acq_params_file

    The computed file passed to the --datain option of topup

  • corrected_se_epi_file

    The .nii.gz image containing all distortion corrected volumes from the two input acquisitions

  • fmap_hz_file

    The fieldmap in hertz (Hz)

  • fmap_rads_file

    The fieldmap in radians per second (rad/s)

  • fmap_mag_file

    The 'magnitude' image (mean image) computed by averaging all volumes in corrected_se_epi_file

  • fmap_mag_brain_file

    The result of applying brain-extraction to fmap_mag_file

  • fmap_mag_brain_mask_file

    The brain mask produced during the brain-extraction of fmap_mag_file

Workflow diagram

Workflow diagram

Contributing

  1. Have or install a recent version of poetry (version >= 1.1)
  2. Fork the repo
  3. Setup a virtual environment (however you prefer)
  4. Run poetry install
  5. Run pre-commit install
  6. Add your changes
  7. Commit your changes + push to your fork
  8. Open a PR

About

Nipype workflow to generate fieldmaps from EPI acquisitions with differing phase-encoding directions

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages