Skip to content

Latest commit

 

History

History
115 lines (81 loc) · 4.04 KB

README.md

File metadata and controls

115 lines (81 loc) · 4.04 KB

Nekton

Python Application TestingTest and Release Python VersionsPackage version

A python package for DICOM to NifTi and NifTi to DICOM-SEG and GSPS conversion

SETUP

The python package is available for use on PyPI. It can be setup simply via pip

pip install nekton

To the check the setup, simply check the version number of the nekton package by

python -c 'import nekton; print(nekton.__version__)'

DICOM to NifTi

The DICOM to NifTi conversion in the package is based on a wrapper around the dcm2niix software.

Usage

from nekton.dcm2nii import Dcm2Nii
converter = Dcm2Nii()
converted_files = converter.run(dicom_directory='/test_files/CT5N',  out_directory='/test_files/CT5N', name='Test')
# Converted 5 DCM to Nifti; Output stored @ /test_files/CT5N
print(converted_files)
# ['/test_files/CT5N/Test_SmartScore_-_Gated_0.5_sec_20010101000000_5.nii.gz']

Parameters converter.run:

  • dicom_directory (Path): path to directory with Dicoms
  • dicom_directory (Path, optional): directory to store the output nifti
  • name (str, optional): Name to be given to the output file. Defaults to "".

Returns:

  • List[Path]: output list of Nifti files

Notes

  • The renaming functionality retains the suffixes from the original program.
  • The BIDS sidecar json is retained as well.

NifTi to DICOM-SEG

The NifTi to DICOM-SEG within nekton converts incoming segmentation NifTi to DICOM-SEG. The matching of the segmentation index to a text label is done via json file using the schema suggested by dcmqi. The json can be generated using the gui also an example can be seen here.

Currently, nekton supports creation of multiclass DICOM-SEG of two types-

  • single layer DICOM-SEG, where each non-empty slice has an individual file
  • multi layer DICOM-SEG, where all the n slices are rolled into a single file

Usage

  1. NifTi to single layer DICOM-SEG
from nekton.nii2dcm import Nii2DcmSeg
import glob
converter = Nii2DcmSeg()
path_dcms = [path for path in glob.glob(dir_dcms)]
path_mapping = "mapping.json"
path_seg_nifti = "CT5N_segmentation.nii.gz"
dcmsegs = converter_dcmseg.multiclass_converter(
        segfile = path_seg_nifti, segMapping= path_mapping, dcmfiles =path_dcms, multiLayer=False
    )
print (len(dcmsegs))
# 3
  1. NifTi to multi layer DICOM-SEG
from nekton.nii2dcm import Nii2DcmSeg
import glob
converter = Nii2DcmSeg()
path_dcms = [path for path in glob.glob(dir_dcms)]
path_mapping = "mapping.json"
path_seg_nifti = "CT5N_segmentation.nii.gz"
dcmsegs = converter.multiclass_converter(
        segfile = path_seg_nifti, segMapping= path_mapping, dcmfiles =path_dcms, multiLayer=True
    )
print (len(dcmsegs))
# 1

Parameters converter.multiclass_converter:

  • segfile (Path): path to the nifti segmentation file
  • segMapping (Path): path to the dcmqii format segmentation mapping json
  • dcmfiles (List[Path]): list of paths of all the source dicom files
  • multiLayer (bool, optional): create a single multilayer dicomseg. Defaults to False.

Returns:

  • List[Path]: list of paths of all generated dicomseg files

Notes

  • Multilabel NifTi(in the form of a NifTi file for a single label) to DICOM-SEG is under development.

NifTi to GSPS

This feature will be available in a future release of the nekton