siemens¶

Module contents¶

Utilities for handling Siemens DICOM data.

Subpackages¶

csa

Submodules¶

dicom_parser.utils.siemens.mosaic module¶

Definition of the Mosaic class that decodes the 2D encoding of 3D volumes used by Siemens. For more information read this NiBabel article or see pages 10-12 in here.

class dicom_parser.utils.siemens.mosaic.Mosaic(mosaic_array: numpy.ndarray, header: dicom_parser.header.Header)¶

Bases: object

A Siemens mosaic of 2D images representing a single volume.

CSA_ASCII_HEADER_KEY: str = 'MrPhoenixProtocol'¶

CSA_ASCII_SLICE_ARRAY_KEY: str = 'sSliceArray'¶

CSA_SERIES_INFO_KEY: str = 'CSASeriesHeaderInfo'¶

fold() → numpy.ndarray¶

Folds the 2D mosaic to become a 3D volume.

Returns: 3D volume
Return type: np.ndarray

get_image_position(iop: numpy.ndarray) → Tuple[float, float, float]¶

Returns a fixed Image Position (Patient) header field value.

Parameters: iop (np.ndarray) – Image Orientation (Patient) header field value

References

https://nipy.org/nibabel/dicom/dicom_mosaic.html

Returns: Image Position (Patient) header field value
Return type: Tuple[float, float, float]

get_image_shape() → Tuple[int, int]¶

Returns the 2D shape of a single image within the mosaic.

Returns: Single image shape
Return type: Tuple[int, int]

get_mosaic_dimensions() → tuple¶

Returns the number of rows and columns that make up the mosaic.

Returns: n_rows, n_columns
Return type: tuple

get_n_images() → int¶

Returns the number of images encoded in the mosaic.

Returns: Number of images encoded in mosaic
Return type: int

get_tile(i_row: int, i_column: int) → numpy.ndarray¶

Cut out a tile of the mosaic by row and column indices.

Parameters

i_row (int) – Row index
i_column (int) – Column index

Returns

A single tile at the (i_row, i_column) position

Return type

np.ndarray

get_tiles() → list¶

Cuts out the tiles (2D slices) from the mosaic.

Returns: Tiles collected by row
Return type: list

get_volume_shape() → Tuple[float, float, float]¶

Returns the dimensions of the volume that will be created.

Returns: Volume shape
Return type: Tuple[float, float, float]

tiles_to_volume(tiles: list) → numpy.ndarray¶

Transforms a list of tiles to a correctly oriented volume.

Parameters: tiles (list) – List of 2D slices as parsed from the mosaic
Returns: Orientation-fixed volume
Return type: np.ndarray

dicom_parser.utils.siemens.private_tags module¶

Siemens specific private tags they may not be accessible by keyword using pydicom.

dicom_parser.utils.siemens.private_tags.B_MATRIX_INDICES = array([0, 1, 2, 1, 3, 4, 2, 4, 5])¶: The raw (6, 1) vector returned represents the upper triangle of a symmetric B matrix.

dicom_parser.utils.siemens.private_tags.b_matrix_to_q_vector(b_matrix: numpy.ndarray, tol: Optional[float] = None) → numpy.ndarray¶

Estimate q-vector from B matrix.

References

https://github.com/nipy/nibabel/blob/ea68c4ecf914e9b0486b2b01c72fce345ba186fb/nibabel/nicom/dwiparams.py#L26

Parameters: b_matrix (np.ndarray (3, 3)) – B matrix
Returns: q-vector
Return type: np.ndarray

dicom_parser.utils.siemens.private_tags.nearest_pos_semi_def(b_matrix: numpy.ndarray)¶

Least squares positive semi-definite tensor estimation.

References

Niethammer M, San Jose Estepar R, Bouix S, Shenton M, Westin CF. On diffusion tensor estimation. Conf Proc IEEE Eng Med Biol Soc. 2006;1:2622-5. PubMed PMID: 17946125; PubMed Central PMCID: PMC2791793.
https://github.com/nipy/nibabel/blob/ea68c4ecf914e9b0486b2b01c72fce345ba186fb/nibabel/nicom/dwiparams.py#L70

Parameters: b_matrix (np.ndrray) – B matrix
Returns: npds – Estimated nearest positive semi-definite array to matrix B
Return type: (3,3) array

Examples

>>> B = np.diag([1, 1, -1])
>>> nearest_pos_semi_def(B)
array([[0.75, 0.  , 0.  ],
       [0.  , 0.75, 0.  ],
       [0.  , 0.  , 0.  ]])

dicom_parser.utils.siemens.private_tags.parse_siemens_b_matrix(value: bytes) → numpy.ndarray¶

Parses the Siemens B matrix header field value.

Parameters: value (bytes) – Raw B matrix header field value
Returns: Parsed B matrix
Return type: np.ndarray

dicom_parser.utils.siemens.private_tags.parse_siemens_bandwith_per_pixel_phase_encode(value: bytes)¶

dicom_parser.utils.siemens.private_tags.parse_siemens_csa_header(value: bytes) → dict¶

dicom_parser.utils.siemens.private_tags.parse_siemens_gradient_direction(value: bytes) → list¶

Parses a SIEMENS MR image’s B-vector as represented in the private (0019, 100E) DiffusionGradientDirection DICOM tag.

Parameters: value (bytes) – SIEMENS private DiffusionGradientDirection data element.
Returns: Gradient directions (B-vector)
Return type: list

dicom_parser.utils.siemens.private_tags.parse_siemens_number_of_slices_in_mosaic(value: bytes) → int¶

dicom_parser.utils.siemens.private_tags.parse_siemens_slice_timing(value: bytes) → list¶

Parses a SIEMENS MR image’s slice timing as saved in the private (0019, 1029) MosaicRefAcqTimes tag to a list of floats representing slice times in milliseconds.

Parameters: value (bytes) – SIEMENS private MosaicRefAcqTimes data element
Returns: Slice times in milliseconds
Return type: list