dosma.DicomReader

class dosma.DicomReader(num_workers: int = 0, verbose: bool = False, group_by: Union[str, int, Sequence[Union[str, int]]] = 'EchoNumbers', sort_by: Optional[Union[str, int, Sequence[Union[str, int]]]] = None, ignore_ext: bool = False, default_ornt: Optional[Tuple[str, str]] = None)[source]

A class for reading DICOM files.

Variables

  • num_workers (int, optional) – Number of workers to use for loading.

  • verbose (bool, optional) – If True, show loading progress bar.

  • group_by (str(s) or int(s), optional) – DICOM attribute(s) used to group dicoms. This can be the attribute tag name (str) or tag number (int).

  • sort_by (str(s) or int(s), optional) – DICOM attribute(s) used to sort dicoms. This sorting is done after sorting files in alphabetical order.

  • ignore_ext (bool, optional) – If True, ignore extension (".dcm") when loading dicoms from directory.

  • default_ornt (Tuple[str, str], optional) – Default in-plane orientation to use if orientation cannot be determined from DICOM header. If not specified and orientation cannot be determined, error will be raised.

  • data_format_code (ImageDataFormat) – The supported image data format.

Examples

>>> # Load single dicom
>>> dr = DicomReader()
>>> mv = dr.load("/path/to/dicom/file", group_by=None)[0]

>>> # Load multi-echo MRI data
>>> dr = DicomReader(num_workers=0, verbose=True)
>>> mvs = dr.load("/dicoms/directory", group_by="EchoTime", sort_by="InstanceNumber")

>>> # Use the same loader for multiple multi-echo time-series MRI scans
>>> dr = DicomReader(group_by=["EchoTime", "TriggerTime", sort_by="InstanceNumber")
>>> scans = [dr.load(dcm_dir) for dcm_dir in ["/dicom/dir1", "/dicom/dir2", "/dicom/dir3"]]
__init__(num_workers: int = 0, verbose: bool = False, group_by: Union[str, int, Sequence[Union[str, int]]] = 'EchoNumbers', sort_by: Optional[Union[str, int, Sequence[Union[str, int]]]] = None, ignore_ext: bool = False, default_ornt: Optional[Tuple[str, str]] = None)[source]
Parameters

  • num_workers (int, optional) – Number of workers to use for loading.

  • verbose (bool, optional) – If True, show loading progress bar.

  • group_by (str(s) or int(s), optional) – DICOM attribute(s) used to group dicoms. This can be the attribute tag name (str) or tag number (int).

  • sort_by (str(s) or int(s), optional) – DICOM attribute(s) used to sort dicoms. This sorting is done after sorting files in alphabetical order.

  • ignore_ext (bool, optional) – If True, ignore extension (".dcm") when loading dicoms from directory.

  • default_ornt (Tuple[str, str], optional) – Default in-plane orientation to use if orientation cannot be determined from DICOM header. If not specified and orientation cannot be determined, error will be raised.

Methods

__init__([num_workers, verbose, group_by, …])

param num_workers

Number of workers to use for loading.

get_files(path[, include, exclude, …])

Get dicom files from directory.

load(path[, group_by, sort_by, ignore_ext, …])

Load dicoms into MedicalVolume``s grouped by ``group_by tag(s).

load_state_dict(state_dict)

Loads the state dictionary into the object.

state_dict()

Returns the state dictionary for an object.

Attributes

data_format_code