torchio.data.image.Image.width() - Code Metrics - Inspection of "Use (H, W, D) for spatial dimensions" - fepegar/torchio - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Pull Request — master (#270)

by Fernando

created 2020-08-17 18:03 UTC

torchio.data.image.Image.width() A

↳ Parent: torchio.data.image

Complexity

Conditions

Size

Total Lines	4
Code Lines	4

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	1
eloc	4
nop	1
dl	0
loc	4
rs	10
c	0
b	0
f	0

import warnings
from pathlib import Path
from typing import Any, Dict, Tuple, Optional, Union, Sequence, List

import torch
import humanize
import numpy as np
import nibabel as nib
import SimpleITK as sitk

from ..utils import (
    nib_to_sitk,
    get_rotation_and_spacing_from_affine,
    get_stem,
    ensure_4d,
)
from ..torchio import (
    TypeData,
    TypePath,
    TypeTripletInt,
    TypeTripletFloat,
    DATA,
    TYPE,
    AFFINE,
    PATH,
    STEM,
    INTENSITY,
    LABEL,
)
from .io import read_image, write_image


PROTECTED_KEYS = DATA, AFFINE, TYPE, PATH, STEM


class Image(dict):
    r"""TorchIO image.

    For information about medical image orientation, check out `NiBabel docs`_,
    the `3D Slicer wiki`_, `Graham Wideman's website`_, `FSL docs`_ or
    `SimpleITK docs`_.

    Args:
        path: Path to a file or sequence of paths to files that can be read by
            :mod:`SimpleITK` or :mod:`nibabel`, or to a directory containing
            DICOM files. If :py:attr:`tensor` is given, the data in
            :py:attr:`path` will not be read. The data is expected to be 2D or
            3D, and may have multiple channels (see :attr:`num_spatial_dims` and
            :attr:`channels_last`). If a sequence of paths is given, data
            will be concatenated on the channel dimension so spatial
            dimensions must match.
        type: Type of image, such as :attr:`torchio.INTENSITY` or
            :attr:`torchio.LABEL`. This will be used by the transforms to
            decide whether to apply an operation, or which interpolation to use
            when resampling. For example, `preprocessing`_ and `augmentation`_
            intensity transforms will only be applied to images with type
            :attr:`torchio.INTENSITY`. Spatial transforms will be applied to
            all types, and nearest neighbor interpolation is always used to
            resample images with type :attr:`torchio.LABEL`.
            The type :attr:`torchio.SAMPLING_MAP` may be used with instances of
            :py:class:`~torchio.data.sampler.weighted.WeightedSampler`.
        tensor: If :py:attr:`path` is not given, :attr:`tensor` must be a 4D
            :py:class:`torch.Tensor` or NumPy array with dimensions
            :math:`(C, H, W, D)`. If it is not 4D, TorchIO will try to guess
            the dimensions meanings. If 2D, the shape will be interpreted as
            :math:`(H, W)`. If 3D, the number of spatial dimensions should be
            determined in :attr:`num_spatial_dims`. If :attr:`num_spatial_dims`
            is not given and the shape is 3 along the first or last dimensions,
            it will be interpreted as a multichannel 2D image. Otherwise, it
            be interpreted as a 3D image with a single channel.
        affine: If :attr:`path` is not given, :attr:`affine` must be a
            :math:`4 \times 4` NumPy array. If ``None``, :attr:`affine` is an
            identity matrix.
        check_nans: If ``True``, issues a warning if NaNs are found
            in the image. If ``False``, images will not be checked for the
            presence of NaNs.
        num_spatial_dims: If ``2`` and the input tensor has 3 dimensions, it
            will be interpreted as a multichannel 2D image. If ``3`` and the
            input has 3 dimensions, it will be interpreted as a
            single-channel 3D volume.
        channels_last: If ``True``, the last dimension of the input will be
            interpreted as the channels. Defaults to ``True`` if :attr:`path` is
            given and ``False`` otherwise.
        **kwargs: Items that will be added to the image dictionary, e.g.
            acquisition parameters.

    TorchIO images are `lazy loaders`_, i.e. the data is only loaded from disk
    when needed.

    Example:
        >>> import torchio
        >>> image = torchio.ScalarImage('t1.nii.gz')  # subclass of Image
        >>> image  # not loaded yet
        ScalarImage(path: t1.nii.gz; type: intensity)
        >>> times_two = 2 * image.data  # data is loaded and cached here
        >>> image
        ScalarImage(shape: (1, 256, 256, 176); spacing: (1.00, 1.00, 1.00); orientation: PIR+; memory: 44.0 MiB; type: intensity)
        >>> image.save('doubled_image.nii.gz')

    .. _lazy loaders: https://en.wikipedia.org/wiki/Lazy_loading
    .. _preprocessing: https://torchio.readthedocs.io/transforms/preprocessing.html#intensity
    .. _augmentation: https://torchio.readthedocs.io/transforms/augmentation.html#intensity
    .. _NiBabel docs: https://nipy.org/nibabel/image_orientation.html
    .. _3D Slicer wiki: https://www.slicer.org/wiki/Coordinate_systems
    .. _FSL docs: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Orientation%20Explained
    .. _SimpleITK docs: https://simpleitk.readthedocs.io/en/master/fundamentalConcepts.html
    .. _Graham Wideman's website: http://www.grahamwideman.com/gw/brain/orientation/orientterms.htm
    """
    def __init__(
            self,
            path: Union[TypePath, Sequence[TypePath], None] = None,
            type: str = None,
            tensor: Optional[TypeData] = None,
            affine: Optional[TypeData] = None,
            check_nans: bool = True,
            num_spatial_dims: Optional[int] = None,
            channels_last: Optional[bool] = None,
            **kwargs: Dict[str, Any],
            ):
        self.check_nans = check_nans
        self.num_spatial_dims = num_spatial_dims

        if type is None:
            warnings.warn(
                'Not specifying the image type is deprecated and will be'
                ' mandatory in the future. You can probably use ScalarImage or'
                ' LabelMap instead'
            )
            type = INTENSITY

        if path is None and tensor is None:
            raise ValueError('A value for path or tensor must be given')
        self._loaded = False

        # Number of channels are typically stored in the last dimensions in disk
        # But if a tensor is given, the channels should be in the first dim
        if channels_last is None:
            channels_last = path is not None
        self.channels_last = channels_last

        tensor = self.parse_tensor(tensor)
        affine = self.parse_affine(affine)
        if tensor is not None:
            self[DATA] = tensor
            self[AFFINE] = affine
            self._loaded = True
        for key in PROTECTED_KEYS:
            if key in kwargs:
                message = f'Key "{key}" is reserved. Use a different one'
                raise ValueError(message)

        super().__init__(**kwargs)
        self.path = self._parse_path(path)
        self[PATH] = '' if self.path is None else str(self.path)
        self[STEM] = '' if self.path is None else get_stem(self.path)
        self[TYPE] = type

    def __repr__(self):
        properties = []
        if self._loaded:
            properties.extend([
                f'shape: {self.shape}',
                f'spacing: {self.get_spacing_string()}',
                f'orientation: {"".join(self.orientation)}+',
                f'memory: {humanize.naturalsize(self.memory, binary=True)}',
            ])
        else:
            properties.append(f'path: "{self.path}"')
        properties.append(f'type: {self.type}')
        properties = '; '.join(properties)
        string = f'{self.__class__.__name__}({properties})'
        return string

    def __getitem__(self, item):
        if item in (DATA, AFFINE):
            if item not in self:
                self.load()
        return super().__getitem__(item)

    def __array__(self):
        return self[DATA].numpy()

    def __copy__(self):
        kwargs = dict(
            tensor=self.data,
            affine=self.affine,
            type=self.type,
            path=self.path,
            channels_last=False,
        )
        for key, value in self.items():
            if key in PROTECTED_KEYS: continue
            kwargs[key] = value  # should I copy? deepcopy?
        return self.__class__(**kwargs)

    @property
    def data(self):
        return self[DATA]

    @property
    def tensor(self):
        return self.data

    @property
    def affine(self):
        return self[AFFINE]

    @property
    def type(self):
        return self[TYPE]

    @property
    def shape(self) -> Tuple[int, int, int, int]:
        return tuple(self.data.shape)

    @property
    def spatial_shape(self) -> TypeTripletInt:
        return self.shape[1:]

    def check_is_2d(self):
        if not self.is_2d():
            message = f'Image is not 2D. Spatial shape: {self.spatial_shape}'
            raise RuntimeError(message)

    @property
    def height(self) -> int:
        self.check_is_2d()
        return self.spatial_shape[0]

    @property
    def width(self) -> int:
        self.check_is_2d()
        return self.spatial_shape[1]

    @property
    def orientation(self):
        return nib.aff2axcodes(self.affine)

    @property
    def spacing(self):
        _, spacing = get_rotation_and_spacing_from_affine(self.affine)
        return tuple(spacing)

    @property
    def memory(self):
        return np.prod(self.shape) * 4  # float32, i.e. 4 bytes per voxel

    def axis_name_to_index(self, axis: str):
        """Convert an axis name to an axis index.

        Args:
            axis: Possible inputs are ``'Left'``, ``'Right'``, ``'Anterior'``,
            ``'Posterior'``, ``'Inferior'``, ``'Superior'``. Lower-case versions
            and first letters are also valid, as only the first letter will be
            used.

        .. note:: If you are working with animals, you should probably use
            ``'Superior'``, ``'Inferior'``, ``'Anterior'`` and ``'Posterior'``
            for ``'Dorsal'``, ``'Ventral'``, ``'Rostral'`` and ``'Caudal'``,
            respectively.
        """
        if not isinstance(axis, str):
            raise ValueError('Axis must be a string')
        axis = axis[0].upper()

        # Generally, TorchIO tensors are (C, H, W, D)
        if axis == 'H':
            return 1
        elif axis == 'W':
            return 2
        else:
            try:
                index = self.orientation.index(axis)
            except ValueError:
                index = self.orientation.index(self.flip_axis(axis))
            # Return negative indices so that it does not matter whether we
            # refer to spatial dimensions or not
            index = -4 + index
            return index

    @staticmethod
    def flip_axis(axis):
        if axis == 'R': return 'L'
        elif axis == 'L': return 'R'
        elif axis == 'A': return 'P'
        elif axis == 'P': return 'A'
        elif axis == 'I': return 'S'
        elif axis == 'S': return 'I'
        else:
            message = (
                f'Axis not understood. Please use a value in {tuple("LRAPIS")}'
            )
            raise ValueError(message)

    def get_spacing_string(self):
        strings = [f'{n:.2f}' for n in self.spacing]
        string = f'({", ".join(strings)})'
        return string

    def get_bounds(self):
        """Get image bounds in mm."""
        first_index = 3 * (-0.5,)
        last_index = np.array(self.spatial_shape) - 0.5
        first_point = nib.affines.apply_affine(self.affine, first_index)
        last_point = nib.affines.apply_affine(self.affine, last_index)
        array = np.array((first_point, last_point))
        bounds_x, bounds_y, bounds_z = array.T.tolist()
        return bounds_x, bounds_y, bounds_z

    @staticmethod
    def _parse_single_path(
            path: TypePath
            ) -> Path:
        try:
            path = Path(path).expanduser()
        except TypeError:
            message = (
                f'Expected type str or Path but found {path} with '
                f'{type(path)} instead'
            )
            raise TypeError(message)
        except RuntimeError:
            message = (
                f'Conversion to path not possible for variable: {path}'
            )
            raise RuntimeError(message)

        if not (path.is_file() or path.is_dir()):   # might be a dir with DICOM
            raise FileNotFoundError(f'File not found: {path}')
        return path

    def _parse_path(
            self,
            path: Union[TypePath, Sequence[TypePath]]
            ) -> Union[Path, List[Path]]:
        if path is None:
            return None
        if isinstance(path, (str, Path)):
            return self._parse_single_path(path)
        else:
            return [self._parse_single_path(p) for p in path]

    def parse_tensor(self, tensor: TypeData) -> torch.Tensor:
        if tensor is None:
            return None
        if isinstance(tensor, np.ndarray):
            tensor = torch.from_numpy(tensor.astype(np.float32))
        elif isinstance(tensor, torch.Tensor):
            tensor = tensor.float()
        tensor = self.parse_tensor_shape(tensor)
        if self.check_nans and torch.isnan(tensor).any():
            warnings.warn(f'NaNs found in tensor')
        return tensor

    def parse_tensor_shape(self, tensor: torch.Tensor) -> torch.Tensor:
        return ensure_4d(tensor, self.channels_last, self.num_spatial_dims)

    @staticmethod
    def parse_affine(affine: np.ndarray) -> np.ndarray:
        if affine is None:
            return np.eye(4)
        if not isinstance(affine, np.ndarray):
            raise TypeError(f'Affine must be a NumPy array, not {type(affine)}')
        if affine.shape != (4, 4):
            raise ValueError(f'Affine shape must be (4, 4), not {affine.shape}')
        return affine

    def load(self) -> None:
        r"""Load the image from disk.

        Returns:
            Tuple containing a 4D tensor of size :math:`(C, H, W, D)` and a 2D
            :math:`4 \times 4` affine matrix to convert voxel indices to world
            coordinates.
        """
        if self._loaded:
            return

        paths = self.path if isinstance(self.path, list) else [self.path]
        tensor, affine = read_image(paths[0])
        tensor = self.parse_tensor_shape(tensor)

        if self.check_nans and torch.isnan(tensor).any():
            warnings.warn(f'NaNs found in file "{paths[0]}"')

        tensors = [tensor]
        for path in paths[1:]:
            new_tensor, new_affine = read_image(path)
            new_tensor = self.parse_tensor_shape(new_tensor)

            if self.check_nans and torch.isnan(tensor).any():
                warnings.warn(f'NaNs found in file "{path}"')

            if not np.array_equal(affine, new_affine):
                message = 'Files have different affine matrices'
                warnings.warn(message, RuntimeWarning)

            if not tensor.shape[1:] == new_tensor.shape[1:]:
                message = (
                    f'Files shape do not match, found {tensor.shape}'
                    f'and {new_tensor.shape}'
                )
                RuntimeError(message)

            tensors.append(new_tensor)

        tensor = torch.cat(tensors)

        self[DATA] = tensor
        self[AFFINE] = affine
        self._loaded = True

    def save(self, path, squeeze=True, channels_last=True):
        """Save image to disk.

        Args:
            path: String or instance of :py:class:`pathlib.Path`.
            squeeze: If ``True``, the singleton dimensions will be removed
                before saving.
            channels_last: If ``True``, the channels will be saved in the last
                dimension.
        """
        write_image(
            self[DATA],
            self[AFFINE],
            path,
            squeeze=squeeze,
            channels_last=channels_last,
        )

    def is_2d(self) -> bool:
        return self.shape[-1] == 1

    def numpy(self) -> np.ndarray:
        """Get a NumPy array containing the image data."""
        return np.asarray(self)

    def as_sitk(self) -> sitk.Image:
        """Get the image as an instance of :py:class:`sitk.Image`."""
        return nib_to_sitk(self[DATA], self[AFFINE])

    def get_center(self, lps: bool = False) -> TypeTripletFloat:
        """Get image center in RAS+ or LPS+ coordinates.

        Args:
            lps: If ``True``, the coordinates will be in LPS+ orientation, i.e.
                the first dimension grows towards the left, etc. Otherwise, the
                coordinates will be in RAS+ orientation.
        """
        size = np.array(self.spatial_shape)
        center_index = (size - 1) / 2
        r, a, s = nib.affines.apply_affine(self.affine, center_index)
        if lps:
            return (-r, -a, s)
        else:
            return (r, a, s)

    def set_check_nans(self, check_nans: bool):
        self.check_nans = check_nans

    def crop(self, index_ini: TypeTripletInt, index_fin: TypeTripletInt):
        new_origin = nib.affines.apply_affine(self.affine, index_ini)
        new_affine = self.affine.copy()
        new_affine[:3, 3] = new_origin
        i0, j0, k0 = index_ini
        i1, j1, k1 = index_fin
        patch = self.data[:, i0:i1, j0:j1, k0:k1].clone()
        kwargs = dict(
            tensor=patch,
            affine=new_affine,
            type=self.type,
            path=self.path,
            channels_last=False,
        )
        for key, value in self.items():
            if key in PROTECTED_KEYS: continue
            kwargs[key] = value  # should I copy? deepcopy?
        return self.__class__(**kwargs)


class ScalarImage(Image):
    """Alias for :py:class:`~torchio.Image` of type :py:attr:`torchio.INTENSITY`.

    Example:
        >>> import torch
        >>> import torchio
        >>> # Loading from a file
        >>> t1_image = torchio.ScalarImage('t1.nii.gz')
        >>> dmri = torchio.ScalarImage(tensor=torch.rand(32, 128, 128, 88))
        >>> image = torchio.ScalarImage('safe_image.nrrd', check_nans=False)
        >>> data, affine = image.data, image.affine
        >>> affine.shape
        (4, 4)
        >>> image.data is image[torchio.DATA]
        True
        >>> image.data is image.tensor
        True
        >>> type(image.data)
        torch.Tensor

    See :py:class:`~torchio.Image` for more information.

    Raises:
        ValueError: A :py:attr:`type` is used for instantiation.
    """
    def __init__(self, *args, **kwargs):
        if 'type' in kwargs and kwargs['type'] != INTENSITY:
            raise ValueError('Type of ScalarImage is always torchio.INTENSITY')
        kwargs.update({'type': INTENSITY})
        super().__init__(*args, **kwargs)


class LabelMap(Image):
    """Alias for :py:class:`~torchio.Image` of type :py:attr:`torchio.LABEL`.

    Example:
        >>> import torch
        >>> import torchio
        >>> labels = torchio.LabelMap(tensor=torch.rand(128, 128, 68) > 0.5)
        >>> labels = torchio.LabelMap('t1_seg.nii.gz')  # loading from a file
        >>> tpm = torchio.LabelMap(                     # loading from files
        ...     'gray_matter.nii.gz',
        ...     'white_matter.nii.gz',
        ...     'csf.nii.gz',
        ... )

    See :py:class:`~torchio.data.image.Image` for more information.

    Raises:
        ValueError: If a value for :py:attr:`type` is given.
    """
    def __init__(self, *args, **kwargs):
        if 'type' in kwargs and kwargs['type'] != LABEL:
            raise ValueError('Type of LabelMap is always torchio.LABEL')
        kwargs.update({'type': LABEL})
        super().__init__(*args, **kwargs)


1			import warnings
2			from pathlib import Path
3			from typing import Any, Dict, Tuple, Optional, Union, Sequence, List
4
5			import torch
6			import humanize
7			import numpy as np
8			import nibabel as nib
9			import SimpleITK as sitk
10
11			from ..utils import (
12			nib_to_sitk,
13			get_rotation_and_spacing_from_affine,
14			get_stem,
15			ensure_4d,
16			)
17			from ..torchio import (
18			TypeData,
19			TypePath,
20			TypeTripletInt,
21			TypeTripletFloat,
22			DATA,
23			TYPE,
24			AFFINE,
25			PATH,
26			STEM,
27			INTENSITY,
28			LABEL,
29			)
30			from .io import read_image, write_image
31
32
33			PROTECTED_KEYS = DATA, AFFINE, TYPE, PATH, STEM
34
35
36			class Image(dict):
37			r"""TorchIO image.
38
39			For information about medical image orientation, check out `NiBabel docs`_,
40			the `3D Slicer wiki`_, `Graham Wideman's website`_, `FSL docs`_ or
41			`SimpleITK docs`_.
42
43			Args:
44			path: Path to a file or sequence of paths to files that can be read by
45			:mod:`SimpleITK` or :mod:`nibabel`, or to a directory containing
46			DICOM files. If :py:attr:`tensor` is given, the data in
47			:py:attr:`path` will not be read. The data is expected to be 2D or
48			3D, and may have multiple channels (see :attr:`num_spatial_dims` and
49			:attr:`channels_last`). If a sequence of paths is given, data
50			will be concatenated on the channel dimension so spatial
51			dimensions must match.
52			type: Type of image, such as :attr:`torchio.INTENSITY` or
53			:attr:`torchio.LABEL`. This will be used by the transforms to
54			decide whether to apply an operation, or which interpolation to use
55			when resampling. For example, `preprocessing`_ and `augmentation`_
56			intensity transforms will only be applied to images with type
57			:attr:`torchio.INTENSITY`. Spatial transforms will be applied to
58			all types, and nearest neighbor interpolation is always used to
59			resample images with type :attr:`torchio.LABEL`.
60			The type :attr:`torchio.SAMPLING_MAP` may be used with instances of
61			:py:class:`~torchio.data.sampler.weighted.WeightedSampler`.
62			tensor: If :py:attr:`path` is not given, :attr:`tensor` must be a 4D
63			:py:class:`torch.Tensor` or NumPy array with dimensions
64			:math:`(C, H, W, D)`. If it is not 4D, TorchIO will try to guess
65			the dimensions meanings. If 2D, the shape will be interpreted as
66			:math:`(H, W)`. If 3D, the number of spatial dimensions should be
67			determined in :attr:`num_spatial_dims`. If :attr:`num_spatial_dims`
68			is not given and the shape is 3 along the first or last dimensions,
69			it will be interpreted as a multichannel 2D image. Otherwise, it
70			be interpreted as a 3D image with a single channel.
71			affine: If :attr:`path` is not given, :attr:`affine` must be a
72			:math:`4 \times 4` NumPy array. If ``None``, :attr:`affine` is an
73			identity matrix.
74			check_nans: If ``True``, issues a warning if NaNs are found
75			in the image. If ``False``, images will not be checked for the
76			presence of NaNs.
77			num_spatial_dims: If ``2`` and the input tensor has 3 dimensions, it
78			will be interpreted as a multichannel 2D image. If ``3`` and the
79			input has 3 dimensions, it will be interpreted as a
80			single-channel 3D volume.
81			channels_last: If ``True``, the last dimension of the input will be
82			interpreted as the channels. Defaults to ``True`` if :attr:`path` is
83			given and ``False`` otherwise.
84			**kwargs: Items that will be added to the image dictionary, e.g.
85			acquisition parameters.
86
87			TorchIO images are `lazy loaders`_, i.e. the data is only loaded from disk
88			when needed.
89
90			Example:
91			>>> import torchio
92			>>> image = torchio.ScalarImage('t1.nii.gz') # subclass of Image
93			>>> image # not loaded yet
94			ScalarImage(path: t1.nii.gz; type: intensity)
95			>>> times_two = 2 * image.data # data is loaded and cached here
96			>>> image
97			ScalarImage(shape: (1, 256, 256, 176); spacing: (1.00, 1.00, 1.00); orientation: PIR+; memory: 44.0 MiB; type: intensity)
98			>>> image.save('doubled_image.nii.gz')
99
100			.. _lazy loaders: https://en.wikipedia.org/wiki/Lazy_loading
101			.. _preprocessing: https://torchio.readthedocs.io/transforms/preprocessing.html#intensity
102			.. _augmentation: https://torchio.readthedocs.io/transforms/augmentation.html#intensity
103			.. _NiBabel docs: https://nipy.org/nibabel/image_orientation.html
104			.. _3D Slicer wiki: https://www.slicer.org/wiki/Coordinate_systems
105			.. _FSL docs: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Orientation%20Explained
106			.. _SimpleITK docs: https://simpleitk.readthedocs.io/en/master/fundamentalConcepts.html
107			.. _Graham Wideman's website: http://www.grahamwideman.com/gw/brain/orientation/orientterms.htm
108			"""
109			def __init__(
110			self,
111			path: Union[TypePath, Sequence[TypePath], None] = None,
112			type: str = None,
113			tensor: Optional[TypeData] = None,
114			affine: Optional[TypeData] = None,
115			check_nans: bool = True,
116			num_spatial_dims: Optional[int] = None,
117			channels_last: Optional[bool] = None,
118			**kwargs: Dict[str, Any],
119			):
120			self.check_nans = check_nans
121			self.num_spatial_dims = num_spatial_dims
122
123			if type is None:
124			warnings.warn(
125			'Not specifying the image type is deprecated and will be'
126			' mandatory in the future. You can probably use ScalarImage or'
127			' LabelMap instead'
128			)
129			type = INTENSITY
130
131			if path is None and tensor is None:
132			raise ValueError('A value for path or tensor must be given')
133			self._loaded = False
134
135			# Number of channels are typically stored in the last dimensions in disk
136			# But if a tensor is given, the channels should be in the first dim
137			if channels_last is None:
138			channels_last = path is not None
139			self.channels_last = channels_last
140
141			tensor = self.parse_tensor(tensor)
142			affine = self.parse_affine(affine)
143			if tensor is not None:
144			self[DATA] = tensor
145			self[AFFINE] = affine
146			self._loaded = True
147			for key in PROTECTED_KEYS:
148			if key in kwargs:
149			message = f'Key "{key}" is reserved. Use a different one'
150			raise ValueError(message)
151
152			super().__init__(**kwargs)
153			self.path = self._parse_path(path)
154			self[PATH] = '' if self.path is None else str(self.path)
155			self[STEM] = '' if self.path is None else get_stem(self.path)
156			self[TYPE] = type
157
158			def __repr__(self):
159			properties = []
160			if self._loaded:
161			properties.extend([
162			f'shape: {self.shape}',
163			f'spacing: {self.get_spacing_string()}',
164			f'orientation: {"".join(self.orientation)}+',
165			f'memory: {humanize.naturalsize(self.memory, binary=True)}',
166			])
167			else:
168			properties.append(f'path: "{self.path}"')
169			properties.append(f'type: {self.type}')
170			properties = '; '.join(properties)
171			string = f'{self.__class__.__name__}({properties})'
172			return string
173
174			def __getitem__(self, item):
175			if item in (DATA, AFFINE):
176			if item not in self:
177			self.load()
178			return super().__getitem__(item)
179
180			def __array__(self):
181			return self[DATA].numpy()
182
183			def __copy__(self):
184			kwargs = dict(
185			tensor=self.data,
186			affine=self.affine,
187			type=self.type,
188			path=self.path,
189			channels_last=False,
190			)
191			for key, value in self.items():
192			if key in PROTECTED_KEYS: continue
193			kwargs[key] = value # should I copy? deepcopy?
194			return self.__class__(**kwargs)
195
196			@property
197			def data(self):
198			return self[DATA]
199
200			@property
201			def tensor(self):
202			return self.data
203
204			@property
205			def affine(self):
206			return self[AFFINE]
207
208			@property
209			def type(self):
210			return self[TYPE]
211
212			@property
213			def shape(self) -> Tuple[int, int, int, int]:
214			return tuple(self.data.shape)
215
216			@property
217			def spatial_shape(self) -> TypeTripletInt:
218			return self.shape[1:]
219
220			def check_is_2d(self):
221			if not self.is_2d():
222			message = f'Image is not 2D. Spatial shape: {self.spatial_shape}'
223			raise RuntimeError(message)
224
225			@property
226			def height(self) -> int:
227			self.check_is_2d()
228			return self.spatial_shape[0]
229
230			@property
231			def width(self) -> int:
232			self.check_is_2d()
233			return self.spatial_shape[1]
234
235			@property
236			def orientation(self):
237			return nib.aff2axcodes(self.affine)
238
239			@property
240			def spacing(self):
241			_, spacing = get_rotation_and_spacing_from_affine(self.affine)
242			return tuple(spacing)
243
244			@property
245			def memory(self):
246			return np.prod(self.shape) * 4 # float32, i.e. 4 bytes per voxel
247
248			def axis_name_to_index(self, axis: str):
249			"""Convert an axis name to an axis index.
250
251			Args:
252			axis: Possible inputs are ``'Left'``, ``'Right'``, ``'Anterior'``,
253			``'Posterior'``, ``'Inferior'``, ``'Superior'``. Lower-case versions
254			and first letters are also valid, as only the first letter will be
255			used.
256
257			.. note:: If you are working with animals, you should probably use
258			``'Superior'``, ``'Inferior'``, ``'Anterior'`` and ``'Posterior'``
259			for ``'Dorsal'``, ``'Ventral'``, ``'Rostral'`` and ``'Caudal'``,
260			respectively.
261			"""
262			if not isinstance(axis, str):
263			raise ValueError('Axis must be a string')
264			axis = axis[0].upper()
265
266			# Generally, TorchIO tensors are (C, H, W, D)
267			if axis == 'H':
268			return 1
269			elif axis == 'W':
270			return 2
271			else:
272			try:
273			index = self.orientation.index(axis)
274			except ValueError:
275			index = self.orientation.index(self.flip_axis(axis))
276			# Return negative indices so that it does not matter whether we
277			# refer to spatial dimensions or not
278			index = -4 + index
279			return index
280
281			@staticmethod
282			def flip_axis(axis):
283			if axis == 'R': return 'L'
284			elif axis == 'L': return 'R'
285			elif axis == 'A': return 'P'
286			elif axis == 'P': return 'A'
287			elif axis == 'I': return 'S'
288			elif axis == 'S': return 'I'
289			else:
290			message = (
291			f'Axis not understood. Please use a value in {tuple("LRAPIS")}'
292			)
293			raise ValueError(message)
294
295			def get_spacing_string(self):
296			strings = [f'{n:.2f}' for n in self.spacing]
297			string = f'({", ".join(strings)})'
298			return string
299
300			def get_bounds(self):
301			"""Get image bounds in mm."""
302			first_index = 3 * (-0.5,)
303			last_index = np.array(self.spatial_shape) - 0.5
304			first_point = nib.affines.apply_affine(self.affine, first_index)
305			last_point = nib.affines.apply_affine(self.affine, last_index)
306			array = np.array((first_point, last_point))
307			bounds_x, bounds_y, bounds_z = array.T.tolist()
308			return bounds_x, bounds_y, bounds_z
309
310			@staticmethod
311			def _parse_single_path(
312			path: TypePath
313			) -> Path:
314			try:
315			path = Path(path).expanduser()
316			except TypeError:
317			message = (
318			f'Expected type str or Path but found {path} with '
319			f'{type(path)} instead'
320			)
321			raise TypeError(message)
322			except RuntimeError:
323			message = (
324			f'Conversion to path not possible for variable: {path}'
325			)
326			raise RuntimeError(message)
327
328			if not (path.is_file() or path.is_dir()): # might be a dir with DICOM
329			raise FileNotFoundError(f'File not found: {path}')
330			return path
331
332			def _parse_path(
333			self,
334			path: Union[TypePath, Sequence[TypePath]]
335			) -> Union[Path, List[Path]]:
336			if path is None:
337			return None
338			if isinstance(path, (str, Path)):
339			return self._parse_single_path(path)
340			else:
341			return [self._parse_single_path(p) for p in path]
342
343			def parse_tensor(self, tensor: TypeData) -> torch.Tensor:
344			if tensor is None:
345			return None
346			if isinstance(tensor, np.ndarray):
347			tensor = torch.from_numpy(tensor.astype(np.float32))
348			elif isinstance(tensor, torch.Tensor):
349			tensor = tensor.float()
350			tensor = self.parse_tensor_shape(tensor)
351			if self.check_nans and torch.isnan(tensor).any():
352			warnings.warn(f'NaNs found in tensor')
353			return tensor
354
355			def parse_tensor_shape(self, tensor: torch.Tensor) -> torch.Tensor:
356			return ensure_4d(tensor, self.channels_last, self.num_spatial_dims)
357
358			@staticmethod
359			def parse_affine(affine: np.ndarray) -> np.ndarray:
360			if affine is None:
361			return np.eye(4)
362			if not isinstance(affine, np.ndarray):
363			raise TypeError(f'Affine must be a NumPy array, not {type(affine)}')
364			if affine.shape != (4, 4):
365			raise ValueError(f'Affine shape must be (4, 4), not {affine.shape}')
366			return affine
367
368			def load(self) -> None:
369			r"""Load the image from disk.
370
371			Returns:
372			Tuple containing a 4D tensor of size :math:`(C, H, W, D)` and a 2D
373			:math:`4 \times 4` affine matrix to convert voxel indices to world
374			coordinates.
375			"""
376			if self._loaded:
377			return
378
379			paths = self.path if isinstance(self.path, list) else [self.path]
380			tensor, affine = read_image(paths[0])
381			tensor = self.parse_tensor_shape(tensor)
382
383			if self.check_nans and torch.isnan(tensor).any():
384			warnings.warn(f'NaNs found in file "{paths[0]}"')
385
386			tensors = [tensor]
387			for path in paths[1:]:
388			new_tensor, new_affine = read_image(path)
389			new_tensor = self.parse_tensor_shape(new_tensor)
390
391			if self.check_nans and torch.isnan(tensor).any():
392			warnings.warn(f'NaNs found in file "{path}"')
393
394			if not np.array_equal(affine, new_affine):
395			message = 'Files have different affine matrices'
396			warnings.warn(message, RuntimeWarning)
397
398			if not tensor.shape[1:] == new_tensor.shape[1:]:
399			message = (
400			f'Files shape do not match, found {tensor.shape}'
401			f'and {new_tensor.shape}'
402			)
403			RuntimeError(message)
404
405			tensors.append(new_tensor)
406
407			tensor = torch.cat(tensors)
408
409			self[DATA] = tensor
410			self[AFFINE] = affine
411			self._loaded = True
412
413			def save(self, path, squeeze=True, channels_last=True):
414			"""Save image to disk.
415
416			Args:
417			path: String or instance of :py:class:`pathlib.Path`.
418			squeeze: If ``True``, the singleton dimensions will be removed
419			before saving.
420			channels_last: If ``True``, the channels will be saved in the last
421			dimension.
422			"""
423			write_image(
424			self[DATA],
425			self[AFFINE],
426			path,
427			squeeze=squeeze,
428			channels_last=channels_last,
429			)
430
431			def is_2d(self) -> bool:
432			return self.shape[-1] == 1
433
434			def numpy(self) -> np.ndarray:
435			"""Get a NumPy array containing the image data."""
436			return np.asarray(self)
437
438			def as_sitk(self) -> sitk.Image:
439			"""Get the image as an instance of :py:class:`sitk.Image`."""
440			return nib_to_sitk(self[DATA], self[AFFINE])
441
442			def get_center(self, lps: bool = False) -> TypeTripletFloat:
443			"""Get image center in RAS+ or LPS+ coordinates.
444
445			Args:
446			lps: If ``True``, the coordinates will be in LPS+ orientation, i.e.
447			the first dimension grows towards the left, etc. Otherwise, the
448			coordinates will be in RAS+ orientation.
449			"""
450			size = np.array(self.spatial_shape)
451			center_index = (size - 1) / 2
452			r, a, s = nib.affines.apply_affine(self.affine, center_index)
453			if lps:
454			return (-r, -a, s)
455			else:
456			return (r, a, s)
457
458			def set_check_nans(self, check_nans: bool):
459			self.check_nans = check_nans
460
461			def crop(self, index_ini: TypeTripletInt, index_fin: TypeTripletInt):
462			new_origin = nib.affines.apply_affine(self.affine, index_ini)
463			new_affine = self.affine.copy()
464			new_affine[:3, 3] = new_origin
465			i0, j0, k0 = index_ini
466			i1, j1, k1 = index_fin
467			patch = self.data[:, i0:i1, j0:j1, k0:k1].clone()
468			kwargs = dict(
469			tensor=patch,
470			affine=new_affine,
471			type=self.type,
472			path=self.path,
473			channels_last=False,
474			)
475			for key, value in self.items():
476			if key in PROTECTED_KEYS: continue
477			kwargs[key] = value # should I copy? deepcopy?
478			return self.__class__(**kwargs)
479
480
481			class ScalarImage(Image):
482			"""Alias for :py:class:`~torchio.Image` of type :py:attr:`torchio.INTENSITY`.
483
484			Example:
485			>>> import torch
486			>>> import torchio
487			>>> # Loading from a file
488			>>> t1_image = torchio.ScalarImage('t1.nii.gz')
489			>>> dmri = torchio.ScalarImage(tensor=torch.rand(32, 128, 128, 88))
490			>>> image = torchio.ScalarImage('safe_image.nrrd', check_nans=False)
491			>>> data, affine = image.data, image.affine
492			>>> affine.shape
493			(4, 4)
494			>>> image.data is image[torchio.DATA]
495			True
496			>>> image.data is image.tensor
497			True
498			>>> type(image.data)
499			torch.Tensor
500
501			See :py:class:`~torchio.Image` for more information.
502
503			Raises:
504			ValueError: A :py:attr:`type` is used for instantiation.
505			"""
506			def __init__(self, args, *kwargs):
507			if 'type' in kwargs and kwargs['type'] != INTENSITY:
508			raise ValueError('Type of ScalarImage is always torchio.INTENSITY')
509			kwargs.update({'type': INTENSITY})
510			super().__init__(args, *kwargs)
511
512
513			class LabelMap(Image):
514			"""Alias for :py:class:`~torchio.Image` of type :py:attr:`torchio.LABEL`.
515
516			Example:
517			>>> import torch
518			>>> import torchio
519			>>> labels = torchio.LabelMap(tensor=torch.rand(128, 128, 68) > 0.5)
520			>>> labels = torchio.LabelMap('t1_seg.nii.gz') # loading from a file
521			>>> tpm = torchio.LabelMap( # loading from files
522			... 'gray_matter.nii.gz',
523			... 'white_matter.nii.gz',
524			... 'csf.nii.gz',
525			... )
526
527			See :py:class:`~torchio.data.image.Image` for more information.
528
529			Raises:
530			ValueError: If a value for :py:attr:`type` is given.
531			"""
532			def __init__(self, args, *kwargs):
533			if 'type' in kwargs and kwargs['type'] != LABEL:
534			raise ValueError('Type of LabelMap is always torchio.LABEL')
535			kwargs.update({'type': LABEL})
536			super().__init__(args, *kwargs)
537

fepegar / torchio

Pull Request — master (#270)

torchio.data.image.Image.width() A

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like