torchio.data.image.Image.__init__() - Code Metrics - Inspection of "Replace assert with try RuntimeError" - fepegar/torchio - Measure and Improve Code Quality continuously with Scrutinizer

Passed

Push — master ( e59e2c...7d4e6f )

by Fernando

created 2021-04-29 18:46 UTC

torchio.data.image.Image.init() C

↳ Parent: torchio.data.image

Complexity

Conditions

Size

Total Lines	44
Code Lines	35

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
eloc	35
dl	0
loc	44
rs	6.6666
c	0
b	0
f	0
cc	9
nop	9

How to fix Many Parameters

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

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

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


PROTECTED_KEYS = DATA, AFFINE, TYPE, PATH, STEM
TypeBound = Tuple[float, float]
TypeBounds = Tuple[TypeBound, TypeBound, TypeBound]

deprecation_message = (
    'Setting the image data with the property setter is deprecated. Use the'
    ' set_data() method instead'
)


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 :attr:`tensor` is given, the data in
            :attr:`path` will not be read.
            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
            :class:`~torchio.data.sampler.weighted.WeightedSampler`.
        tensor: If :attr:`path` is not given, :attr:`tensor` must be a 4D
            :class:`torch.Tensor` or NumPy array with dimensions
            :math:`(C, W, H, D)`.
        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.
        reader: Callable object that takes a path and returns a 4D tensor and a
            2D, :math:`4 \times 4` affine matrix. This can be used if your data
            is saved in a custom format, such as ``.npy`` (see example below).
            If the affine matrix is ``None``, an identity matrix will be used.
        **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 as tio
        >>> import numpy as np
        >>> image = tio.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')
        >>> numpy_reader = lambda path: np.load(path), np.eye(4)
        >>> image = tio.ScalarImage('t1.npy', reader=numpy_reader)

    .. _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 = False,  # removed by ITK by default
            channels_last: bool = False,
            reader: Callable = read_image,
            **kwargs: Dict[str, Any],
            ):
        self.check_nans = check_nans
        self.channels_last = channels_last
        self.reader = reader

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

        tensor = self._parse_tensor(tensor)
        affine = self._parse_affine(affine)
        if tensor is not None:
            self.set_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}"')
        if self._loaded:
            properties.append(f'dtype: {self.data.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,
        )
        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) -> torch.Tensor:
        """Tensor data. Same as :class:`Image.tensor`."""
        return self[DATA]

    @data.setter
    @deprecated(version='0.18.16', reason=deprecation_message)
    def data(self, tensor: TypeData):
        self.set_data(tensor)

    def set_data(self, tensor: TypeData):
        """Store a 4D tensor in the :attr:`data` key and attribute.

        Args:
            tensor: 4D tensor with dimensions :math:`(C, W, H, D)`.
        """
        self[DATA] = self._parse_tensor(tensor, none_ok=False)

    @property
    def tensor(self) -> torch.Tensor:
        """Tensor data. Same as :class:`Image.data`."""
        return self.data

    @property
    def affine(self) -> np.ndarray:
        """Affine matrix to transform voxel indices into world coordinates."""
        return self[AFFINE]

    @affine.setter
    def affine(self, matrix):
        self[AFFINE] = self._parse_affine(matrix)

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

    @property
    def shape(self) -> Tuple[int, int, int, int]:
        """Tensor shape as :math:`(C, W, H, D)`."""
        return tuple(self.data.shape)

    @property
    def spatial_shape(self) -> TypeTripletInt:
        """Tensor spatial shape as :math:`(W, H, D)`."""
        return self.shape[1:]

    def check_is_2d(self) -> None:
        if not self.is_2d():
            message = f'Image is not 2D. Spatial shape: {self.spatial_shape}'
            raise RuntimeError(message)

    @property
    def height(self) -> int:
        """Image height, if 2D."""
        self.check_is_2d()
        return self.spatial_shape[1]

    @property
    def width(self) -> int:
        """Image width, if 2D."""
        self.check_is_2d()
        return self.spatial_shape[0]

    @property
    def orientation(self) -> Tuple[str, str, str]:
        """Orientation codes."""
        return nib.aff2axcodes(self.affine)

    @property
    def spacing(self) -> Tuple[float, float, float]:
        """Voxel spacing in mm."""
        _, spacing = get_rotation_and_spacing_from_affine(self.affine)
        return tuple(spacing)

    @property
    def itemsize(self):
        """Element size of the data type."""
        return self.data.element_size()

    @property
    def memory(self) -> float:
        """Number of Bytes that the tensor takes in the RAM."""
        return np.prod(self.shape) * self.itemsize

    @property
    def bounds(self) -> np.ndarray:
        """Position of centers of voxels in smallest and largest coordinates."""
        ini = 0, 0, 0
        fin = np.array(self.spatial_shape) - 1
        point_ini = nib.affines.apply_affine(self.affine, ini)
        point_fin = nib.affines.apply_affine(self.affine, fin)
        return np.array((point_ini, point_fin))

    @property
    def num_channels(self) -> int:
        """Get the number of channels in the associated 4D tensor."""
        return len(self.data)

    def axis_name_to_index(self, axis: str) -> int:
        """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.

        .. note:: If your images are 2D, you can use ``'Top'``, ``'Bottom'``,
            ``'Left'`` and ``'Right'``.
        """
        # Top and bottom are used for the vertical 2D axis as the use of
        # Height vs Horizontal might be ambiguous

        if not isinstance(axis, str):
            raise ValueError('Axis must be a string')
        axis = axis[0].upper()

        # Generally, TorchIO tensors are (C, W, H, D)
        if axis in 'TB':  # Top, Bottom
            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 = -3 + index
            return index

    # flake8: noqa: E701
    @staticmethod
    def flip_axis(axis: str) -> str:
        if axis == 'R': flipped_axis = 'L'
        elif axis == 'L': flipped_axis = 'R'
        elif axis == 'A': flipped_axis = 'P'
        elif axis == 'P': flipped_axis = 'A'
        elif axis == 'I': flipped_axis = 'S'
        elif axis == 'S': flipped_axis = 'I'
        elif axis == 'T': flipped_axis = 'B'
        elif axis == 'B': flipped_axis = 'T'
        else:
            values = ', '.join('LRPAISTB')
            message = f'Axis not understood. Please use one of: {values}'
            raise ValueError(message)
        return flipped_axis

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

    def get_bounds(self) -> TypeBounds:
        """Get minimum and maximum world coordinates occupied by the image."""
        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 type'
                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]]
            ) -> Optional[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,
            none_ok: bool = True,
            ) -> torch.Tensor:
        if tensor is None:
            if none_ok:
                return None
            else:
                raise RuntimeError('Input tensor cannot be None')
        if isinstance(tensor, np.ndarray):
            tensor = check_uint_to_int(tensor)
            tensor = torch.as_tensor(tensor)
        elif not isinstance(tensor, torch.Tensor):
            message = 'Input tensor must be a PyTorch tensor or NumPy array'
            raise TypeError(message)
        ndim = tensor.ndim
        if ndim != 4:
            raise ValueError(f'Input tensor must be 4D, but it is {ndim}D')
        if tensor.dtype == torch.bool:
            tensor = tensor.to(torch.uint8)
        if self.check_nans and torch.isnan(tensor).any():
            warnings.warn(f'NaNs found in tensor', RuntimeWarning)
        return tensor

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

    @staticmethod
    def _parse_affine(affine: TypeData) -> np.ndarray:
        if affine is None:
            return np.eye(4)
        if isinstance(affine, torch.Tensor):
            affine = affine.numpy()
        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, W, H, 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 = self.read_and_check(paths[0])
        tensors = [tensor]
        for path in paths[1:]:
            new_tensor, new_affine = self.read_and_check(path)
            if not np.array_equal(affine, new_affine):
                message = (
                    'Files have different affine matrices.'
                    f'\nMatrix of {paths[0]}:'
                    f'\n{affine}'
                    f'\nMatrix of {path}:'
                    f'\n{new_affine}'
                )
                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.set_data(tensor)
        self.affine = affine
        self._loaded = True

    def read_and_check(self, path: TypePath) -> Tuple[torch.Tensor, np.ndarray]:
        tensor, affine = self.reader(path)
        tensor = self.parse_tensor_shape(tensor)
        tensor = self._parse_tensor(tensor)
        affine = self._parse_affine(affine)
        if self.channels_last:
            tensor = tensor.permute(3, 0, 1, 2)
        if self.check_nans and torch.isnan(tensor).any():
            warnings.warn(f'NaNs found in file "{path}"', RuntimeWarning)
        return tensor, affine

    def save(self, path: TypePath, squeeze: bool = True) -> None:
        """Save image to disk.

        Args:
            path: String or instance of :class:`pathlib.Path`.
            squeeze: If ``True``, singleton dimensions will be removed
                before saving.
        """
        write_image(
            self.data,
            self.affine,
            path,
            squeeze=squeeze,
        )

    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, **kwargs) -> sitk.Image:
        """Get the image as an instance of :class:`sitk.Image`."""
        return nib_to_sitk(self.data, self.affine, **kwargs)

    def as_pil(self, transpose=True):
        """Get the image as an instance of :class:`PIL.Image`.

        .. note:: Values will be clamped to 0-255 and cast to uint8.
        .. note:: To use this method, `Pillow` needs to be installed:
            `pip install Pillow`.
        """
        try:
            from PIL import Image as ImagePIL
        except ModuleNotFoundError as e:
            message = (
                'Please install Pillow to use Image.as_pil():'
                ' pip install Pillow'
            )
            raise RuntimeError(message) from e

        self.check_is_2d()
        tensor = self.data
        if len(tensor) == 1:
            tensor = torch.cat(3 * [tensor])
        if len(tensor) != 3:
            raise RuntimeError('The image must have 1 or 3 channels')
        if transpose:
            tensor = tensor.permute(3, 2, 1, 0)
        else:
            tensor = tensor.permute(3, 1, 2, 0)
        array = tensor.clamp(0, 255).numpy()[0]
        return ImagePIL.fromarray(array.astype(np.uint8))

    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) -> None:
        self.check_nans = check_nans

    def plot(self, **kwargs) -> None:
        """Plot image."""
        if self.is_2d():
            self.as_pil().show()
        else:
            from ..visualization import plot_volume  # avoid circular import
            plot_volume(self, **kwargs)


class ScalarImage(Image):
    """Image whose pixel values represent scalars.

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

    See :class:`~torchio.Image` for more information.
    """
    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):
    """Image whose pixel values represent categorical labels.

    Intensity transforms are not applied to these images.

    Example:
        >>> import torch
        >>> import torchio as tio
        >>> labels = tio.LabelMap(tensor=torch.rand(1, 128, 128, 68) > 0.5)
        >>> labels = tio.LabelMap('t1_seg.nii.gz')  # loading from a file
        >>> tpm = tio.LabelMap(                     # loading from files
        ...     'gray_matter.nii.gz',
        ...     'white_matter.nii.gz',
        ...     'csf.nii.gz',
        ... )

    See :class:`~torchio.Image` for more information.
    """
    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, Callable
4
5			import torch
6			import humanize
7			import numpy as np
8			import nibabel as nib
9			import SimpleITK as sitk
10			from deprecated import deprecated
11
12			from ..utils import get_stem
13			from ..typing import TypeData, TypePath, TypeTripletInt, TypeTripletFloat
14			from ..constants import DATA, TYPE, AFFINE, PATH, STEM, INTENSITY, LABEL
15			from .io import (
16			ensure_4d,
17			read_image,
18			write_image,
19			nib_to_sitk,
20			check_uint_to_int,
21			get_rotation_and_spacing_from_affine,
22			)
23
24
25			PROTECTED_KEYS = DATA, AFFINE, TYPE, PATH, STEM
26			TypeBound = Tuple[float, float]
27			TypeBounds = Tuple[TypeBound, TypeBound, TypeBound]
28
29			deprecation_message = (
30			'Setting the image data with the property setter is deprecated. Use the'
31			' set_data() method instead'
32			)
33
34
35			class Image(dict):
36			r"""TorchIO image.
37
38			For information about medical image orientation, check out `NiBabel docs`_,
39			the `3D Slicer wiki`_, `Graham Wideman's website`_, `FSL docs`_ or
40			`SimpleITK docs`_.
41
42			Args:
43			path: Path to a file or sequence of paths to files that can be read by
44			:mod:`SimpleITK` or :mod:`nibabel`, or to a directory containing
45			DICOM files. If :attr:`tensor` is given, the data in
46			:attr:`path` will not be read.
47			If a sequence of paths is given, data
48			will be concatenated on the channel dimension so spatial
49			dimensions must match.
50			type: Type of image, such as :attr:`torchio.INTENSITY` or
51			:attr:`torchio.LABEL`. This will be used by the transforms to
52			decide whether to apply an operation, or which interpolation to use
53			when resampling. For example, `preprocessing`_ and `augmentation`_
54			intensity transforms will only be applied to images with type
55			:attr:`torchio.INTENSITY`. Spatial transforms will be applied to
56			all types, and nearest neighbor interpolation is always used to
57			resample images with type :attr:`torchio.LABEL`.
58			The type :attr:`torchio.SAMPLING_MAP` may be used with instances of
59			:class:`~torchio.data.sampler.weighted.WeightedSampler`.
60			tensor: If :attr:`path` is not given, :attr:`tensor` must be a 4D
61			:class:`torch.Tensor` or NumPy array with dimensions
62			:math:`(C, W, H, D)`.
63			affine: If :attr:`path` is not given, :attr:`affine` must be a
64			:math:`4 \times 4` NumPy array. If ``None``, :attr:`affine` is an
65			identity matrix.
66			check_nans: If ``True``, issues a warning if NaNs are found
67			in the image. If ``False``, images will not be checked for the
68			presence of NaNs.
69			reader: Callable object that takes a path and returns a 4D tensor and a
70			2D, :math:`4 \times 4` affine matrix. This can be used if your data
71			is saved in a custom format, such as ``.npy`` (see example below).
72			If the affine matrix is ``None``, an identity matrix will be used.
73			**kwargs: Items that will be added to the image dictionary, e.g.
74			acquisition parameters.
75
76			TorchIO images are `lazy loaders`_, i.e. the data is only loaded from disk
77			when needed.
78
79			Example:
80			>>> import torchio as tio
81			>>> import numpy as np
82			>>> image = tio.ScalarImage('t1.nii.gz') # subclass of Image
83			>>> image # not loaded yet
84			ScalarImage(path: t1.nii.gz; type: intensity)
85			>>> times_two = 2 * image.data # data is loaded and cached here
86			>>> image
87			ScalarImage(shape: (1, 256, 256, 176); spacing: (1.00, 1.00, 1.00); orientation: PIR+; memory: 44.0 MiB; type: intensity)
88			>>> image.save('doubled_image.nii.gz')
89			>>> numpy_reader = lambda path: np.load(path), np.eye(4)
90			>>> image = tio.ScalarImage('t1.npy', reader=numpy_reader)
91
92			.. _lazy loaders: https://en.wikipedia.org/wiki/Lazy_loading
93			.. _preprocessing: https://torchio.readthedocs.io/transforms/preprocessing.html#intensity
94			.. _augmentation: https://torchio.readthedocs.io/transforms/augmentation.html#intensity
95			.. _NiBabel docs: https://nipy.org/nibabel/image_orientation.html
96			.. _3D Slicer wiki: https://www.slicer.org/wiki/Coordinate_systems
97			.. _FSL docs: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Orientation%20Explained
98			.. _SimpleITK docs: https://simpleitk.readthedocs.io/en/master/fundamentalConcepts.html
99			.. _Graham Wideman's website: http://www.grahamwideman.com/gw/brain/orientation/orientterms.htm
100			"""
101			def __init__(
102			self,
103			path: Union[TypePath, Sequence[TypePath], None] = None,
104			type: str = None,
105			tensor: Optional[TypeData] = None,
106			affine: Optional[TypeData] = None,
107			check_nans: bool = False, # removed by ITK by default
108			channels_last: bool = False,
109			reader: Callable = read_image,
110			**kwargs: Dict[str, Any],
111			):
112			self.check_nans = check_nans
113			self.channels_last = channels_last
114			self.reader = reader
115
116			if type is None:
117			warnings.warn(
118			'Not specifying the image type is deprecated and will be'
119			' mandatory in the future. You can probably use tio.ScalarImage'
120			' or tio.LabelMap instead',
121			)
122			type = INTENSITY
123
124			if path is None and tensor is None:
125			raise ValueError('A value for path or tensor must be given')
126			self._loaded = False
127
128			tensor = self._parse_tensor(tensor)
129			affine = self._parse_affine(affine)
130			if tensor is not None:
131			self.set_data(tensor)
132			self.affine = affine
133			self._loaded = True
134			for key in PROTECTED_KEYS:
135			if key in kwargs:
136			message = f'Key "{key}" is reserved. Use a different one'
137			raise ValueError(message)
138
139			super().__init__(**kwargs)
140			self.path = self._parse_path(path)
141
142			self[PATH] = '' if self.path is None else str(self.path)
143			self[STEM] = '' if self.path is None else get_stem(self.path)
144			self[TYPE] = type
145
146			def __repr__(self):
147			properties = []
148			if self._loaded:
149			properties.extend([
150			f'shape: {self.shape}',
151			f'spacing: {self.get_spacing_string()}',
152			f'orientation: {"".join(self.orientation)}+',
153			f'memory: {humanize.naturalsize(self.memory, binary=True)}',
154			])
155			else:
156			properties.append(f'path: "{self.path}"')
157			if self._loaded:
158			properties.append(f'dtype: {self.data.type()}')
159			properties = '; '.join(properties)
160			string = f'{self.__class__.__name__}({properties})'
161			return string
162
163			def __getitem__(self, item):
164			if item in (DATA, AFFINE):
165			if item not in self:
166			self.load()
167			return super().__getitem__(item)
168
169			def __array__(self):
170			return self.data.numpy()
171
172			def __copy__(self):
173			kwargs = dict(
174			tensor=self.data,
175			affine=self.affine,
176			type=self.type,
177			path=self.path,
178			)
179			for key, value in self.items():
180			if key in PROTECTED_KEYS: continue
181			kwargs[key] = value # should I copy? deepcopy?
182			return self.__class__(**kwargs)
183
184			@property
185			def data(self) -> torch.Tensor:
186			"""Tensor data. Same as :class:`Image.tensor`."""
187			return self[DATA]
188
189			@data.setter
190			@deprecated(version='0.18.16', reason=deprecation_message)
191			def data(self, tensor: TypeData):
192			self.set_data(tensor)
193
194			def set_data(self, tensor: TypeData):
195			"""Store a 4D tensor in the :attr:`data` key and attribute.
196
197			Args:
198			tensor: 4D tensor with dimensions :math:`(C, W, H, D)`.
199			"""
200			self[DATA] = self._parse_tensor(tensor, none_ok=False)
201
202			@property
203			def tensor(self) -> torch.Tensor:
204			"""Tensor data. Same as :class:`Image.data`."""
205			return self.data
206
207			@property
208			def affine(self) -> np.ndarray:
209			"""Affine matrix to transform voxel indices into world coordinates."""
210			return self[AFFINE]
211
212			@affine.setter
213			def affine(self, matrix):
214			self[AFFINE] = self._parse_affine(matrix)
215
216			@property
217			def type(self) -> str:
218			return self[TYPE]
219
220			@property
221			def shape(self) -> Tuple[int, int, int, int]:
222			"""Tensor shape as :math:`(C, W, H, D)`."""
223			return tuple(self.data.shape)
224
225			@property
226			def spatial_shape(self) -> TypeTripletInt:
227			"""Tensor spatial shape as :math:`(W, H, D)`."""
228			return self.shape[1:]
229
230			def check_is_2d(self) -> None:
231			if not self.is_2d():
232			message = f'Image is not 2D. Spatial shape: {self.spatial_shape}'
233			raise RuntimeError(message)
234
235			@property
236			def height(self) -> int:
237			"""Image height, if 2D."""
238			self.check_is_2d()
239			return self.spatial_shape[1]
240
241			@property
242			def width(self) -> int:
243			"""Image width, if 2D."""
244			self.check_is_2d()
245			return self.spatial_shape[0]
246
247			@property
248			def orientation(self) -> Tuple[str, str, str]:
249			"""Orientation codes."""
250			return nib.aff2axcodes(self.affine)
251
252			@property
253			def spacing(self) -> Tuple[float, float, float]:
254			"""Voxel spacing in mm."""
255			_, spacing = get_rotation_and_spacing_from_affine(self.affine)
256			return tuple(spacing)
257
258			@property
259			def itemsize(self):
260			"""Element size of the data type."""
261			return self.data.element_size()
262
263			@property
264			def memory(self) -> float:
265			"""Number of Bytes that the tensor takes in the RAM."""
266			return np.prod(self.shape) * self.itemsize
267
268			@property
269			def bounds(self) -> np.ndarray:
270			"""Position of centers of voxels in smallest and largest coordinates."""
271			ini = 0, 0, 0
272			fin = np.array(self.spatial_shape) - 1
273			point_ini = nib.affines.apply_affine(self.affine, ini)
274			point_fin = nib.affines.apply_affine(self.affine, fin)
275			return np.array((point_ini, point_fin))
276
277			@property
278			def num_channels(self) -> int:
279			"""Get the number of channels in the associated 4D tensor."""
280			return len(self.data)
281
282			def axis_name_to_index(self, axis: str) -> int:
283			"""Convert an axis name to an axis index.
284
285			Args:
286			axis: Possible inputs are ``'Left'``, ``'Right'``, ``'Anterior'``,
287			``'Posterior'``, ``'Inferior'``, ``'Superior'``. Lower-case
288			versions and first letters are also valid, as only the first
289			letter will be used.
290
291			.. note:: If you are working with animals, you should probably use
292			``'Superior'``, ``'Inferior'``, ``'Anterior'`` and ``'Posterior'``
293			for ``'Dorsal'``, ``'Ventral'``, ``'Rostral'`` and ``'Caudal'``,
294			respectively.
295
296			.. note:: If your images are 2D, you can use ``'Top'``, ``'Bottom'``,
297			``'Left'`` and ``'Right'``.
298			"""
299			# Top and bottom are used for the vertical 2D axis as the use of
300			# Height vs Horizontal might be ambiguous
301
302			if not isinstance(axis, str):
303			raise ValueError('Axis must be a string')
304			axis = axis[0].upper()
305
306			# Generally, TorchIO tensors are (C, W, H, D)
307			if axis in 'TB': # Top, Bottom
308			return -2
309			else:
310			try:
311			index = self.orientation.index(axis)
312			except ValueError:
313			index = self.orientation.index(self.flip_axis(axis))
314			# Return negative indices so that it does not matter whether we
315			# refer to spatial dimensions or not
316			index = -3 + index
317			return index
318
319			# flake8: noqa: E701
320			@staticmethod
321			def flip_axis(axis: str) -> str:
322			if axis == 'R': flipped_axis = 'L'
323			elif axis == 'L': flipped_axis = 'R'
324			elif axis == 'A': flipped_axis = 'P'
325			elif axis == 'P': flipped_axis = 'A'
326			elif axis == 'I': flipped_axis = 'S'
327			elif axis == 'S': flipped_axis = 'I'
328			elif axis == 'T': flipped_axis = 'B'
329			elif axis == 'B': flipped_axis = 'T'
330			else:
331			values = ', '.join('LRPAISTB')
332			message = f'Axis not understood. Please use one of: {values}'
333			raise ValueError(message)
334			return flipped_axis
335
336			def get_spacing_string(self) -> str:
337			strings = [f'{n:.2f}' for n in self.spacing]
338			string = f'({", ".join(strings)})'
339			return string
340
341			def get_bounds(self) -> TypeBounds:
342			"""Get minimum and maximum world coordinates occupied by the image."""
343			first_index = 3 * (-0.5,)
344			last_index = np.array(self.spatial_shape) - 0.5
345			first_point = nib.affines.apply_affine(self.affine, first_index)
346			last_point = nib.affines.apply_affine(self.affine, last_index)
347			array = np.array((first_point, last_point))
348			bounds_x, bounds_y, bounds_z = array.T.tolist()
349			return bounds_x, bounds_y, bounds_z
350
351			@staticmethod
352			def _parse_single_path(
353			path: TypePath
354			) -> Path:
355			try:
356			path = Path(path).expanduser()
357			except TypeError:
358			message = (
359			f'Expected type str or Path but found {path} with type'
360			f' {type(path)} instead'
361			)
362			raise TypeError(message)
363			except RuntimeError:
364			message = (
365			f'Conversion to path not possible for variable: {path}'
366			)
367			raise RuntimeError(message)
368
369			if not (path.is_file() or path.is_dir()): # might be a dir with DICOM
370			raise FileNotFoundError(f'File not found: "{path}"')
371			return path
372
373			def _parse_path(
374			self,
375			path: Union[TypePath, Sequence[TypePath]]
376			) -> Optional[Union[Path, List[Path]]]:
377			if path is None:
378			return None
379			if isinstance(path, (str, Path)):
380			return self._parse_single_path(path)
381			else:
382			return [self._parse_single_path(p) for p in path]
383
384			def _parse_tensor(
385			self,
386			tensor: TypeData,
387			none_ok: bool = True,
388			) -> torch.Tensor:
389			if tensor is None:
390			if none_ok:
391			return None
392			else:
393			raise RuntimeError('Input tensor cannot be None')
394			if isinstance(tensor, np.ndarray):
395			tensor = check_uint_to_int(tensor)
396			tensor = torch.as_tensor(tensor)
397			elif not isinstance(tensor, torch.Tensor):
398			message = 'Input tensor must be a PyTorch tensor or NumPy array'
399			raise TypeError(message)
400			ndim = tensor.ndim
401			if ndim != 4:
402			raise ValueError(f'Input tensor must be 4D, but it is {ndim}D')
403			if tensor.dtype == torch.bool:
404			tensor = tensor.to(torch.uint8)
405			if self.check_nans and torch.isnan(tensor).any():
406			warnings.warn(f'NaNs found in tensor', RuntimeWarning)
407			return tensor
408
409			def parse_tensor_shape(self, tensor: torch.Tensor) -> torch.Tensor:
410			return ensure_4d(tensor)
411
412			@staticmethod
413			def _parse_affine(affine: TypeData) -> np.ndarray:
414			if affine is None:
415			return np.eye(4)
416			if isinstance(affine, torch.Tensor):
417			affine = affine.numpy()
418			if not isinstance(affine, np.ndarray):
419			raise TypeError(f'Affine must be a NumPy array, not {type(affine)}')
420			if affine.shape != (4, 4):
421			raise ValueError(f'Affine shape must be (4, 4), not {affine.shape}')
422			return affine
423
424			def load(self) -> None:
425			r"""Load the image from disk.
426
427			Returns:
428			Tuple containing a 4D tensor of size :math:`(C, W, H, D)` and a 2D
429			:math:`4 \times 4` affine matrix to convert voxel indices to world
430			coordinates.
431			"""
432			if self._loaded:
433			return
434			paths = self.path if isinstance(self.path, list) else [self.path]
435			tensor, affine = self.read_and_check(paths[0])
436			tensors = [tensor]
437			for path in paths[1:]:
438			new_tensor, new_affine = self.read_and_check(path)
439			if not np.array_equal(affine, new_affine):
440			message = (
441			'Files have different affine matrices.'
442			f'\nMatrix of {paths[0]}:'
443			f'\n{affine}'
444			f'\nMatrix of {path}:'
445			f'\n{new_affine}'
446			)
447			warnings.warn(message, RuntimeWarning)
448			if not tensor.shape[1:] == new_tensor.shape[1:]:
449			message = (
450			f'Files shape do not match, found {tensor.shape}'
451			f'and {new_tensor.shape}'
452			)
453			RuntimeError(message)
454			tensors.append(new_tensor)
455			tensor = torch.cat(tensors)
456			self.set_data(tensor)
457			self.affine = affine
458			self._loaded = True
459
460			def read_and_check(self, path: TypePath) -> Tuple[torch.Tensor, np.ndarray]:
461			tensor, affine = self.reader(path)
462			tensor = self.parse_tensor_shape(tensor)
463			tensor = self._parse_tensor(tensor)
464			affine = self._parse_affine(affine)
465			if self.channels_last:
466			tensor = tensor.permute(3, 0, 1, 2)
467			if self.check_nans and torch.isnan(tensor).any():
468			warnings.warn(f'NaNs found in file "{path}"', RuntimeWarning)
469			return tensor, affine
470
471			def save(self, path: TypePath, squeeze: bool = True) -> None:
472			"""Save image to disk.
473
474			Args:
475			path: String or instance of :class:`pathlib.Path`.
476			squeeze: If ``True``, singleton dimensions will be removed
477			before saving.
478			"""
479			write_image(
480			self.data,
481			self.affine,
482			path,
483			squeeze=squeeze,
484			)
485
486			def is_2d(self) -> bool:
487			return self.shape[-1] == 1
488
489			def numpy(self) -> np.ndarray:
490			"""Get a NumPy array containing the image data."""
491			return np.asarray(self)
492
493			def as_sitk(self, **kwargs) -> sitk.Image:
494			"""Get the image as an instance of :class:`sitk.Image`."""
495			return nib_to_sitk(self.data, self.affine, **kwargs)
496
497			def as_pil(self, transpose=True):
498			"""Get the image as an instance of :class:`PIL.Image`.
499
500			.. note:: Values will be clamped to 0-255 and cast to uint8.
501			.. note:: To use this method, `Pillow` needs to be installed:
502			`pip install Pillow`.
503			"""
504			try:
505			from PIL import Image as ImagePIL
506			except ModuleNotFoundError as e:
507			message = (
508			'Please install Pillow to use Image.as_pil():'
509			' pip install Pillow'
510			)
511			raise RuntimeError(message) from e
512
513			self.check_is_2d()
514			tensor = self.data
515			if len(tensor) == 1:
516			tensor = torch.cat(3 * [tensor])
517			if len(tensor) != 3:
518			raise RuntimeError('The image must have 1 or 3 channels')
519			if transpose:
520			tensor = tensor.permute(3, 2, 1, 0)
521			else:
522			tensor = tensor.permute(3, 1, 2, 0)
523			array = tensor.clamp(0, 255).numpy()[0]
524			return ImagePIL.fromarray(array.astype(np.uint8))
525
526			def get_center(self, lps: bool = False) -> TypeTripletFloat:
527			"""Get image center in RAS+ or LPS+ coordinates.
528
529			Args:
530			lps: If ``True``, the coordinates will be in LPS+ orientation, i.e.
531			the first dimension grows towards the left, etc. Otherwise, the
532			coordinates will be in RAS+ orientation.
533			"""
534			size = np.array(self.spatial_shape)
535			center_index = (size - 1) / 2
536			r, a, s = nib.affines.apply_affine(self.affine, center_index)
537			if lps:
538			return (-r, -a, s)
539			else:
540			return (r, a, s)
541
542			def set_check_nans(self, check_nans: bool) -> None:
543			self.check_nans = check_nans
544
545			def plot(self, **kwargs) -> None:
546			"""Plot image."""
547			if self.is_2d():
548			self.as_pil().show()
549			else:
550			from ..visualization import plot_volume # avoid circular import
551			plot_volume(self, **kwargs)
552
553
554			class ScalarImage(Image):
555			"""Image whose pixel values represent scalars.
556
557			Example:
558			>>> import torch
559			>>> import torchio as tio
560			>>> # Loading from a file
561			>>> t1_image = tio.ScalarImage('t1.nii.gz')
562			>>> dmri = tio.ScalarImage(tensor=torch.rand(32, 128, 128, 88))
563			>>> image = tio.ScalarImage('safe_image.nrrd', check_nans=False)
564			>>> data, affine = image.data, image.affine
565			>>> affine.shape
566			(4, 4)
567			>>> image.data is image[tio.DATA]
568			True
569			>>> image.data is image.tensor
570			True
571			>>> type(image.data)
572			torch.Tensor
573
574			See :class:`~torchio.Image` for more information.
575			"""
576			def __init__(self, args, *kwargs):
577			if 'type' in kwargs and kwargs['type'] != INTENSITY:
578			raise ValueError('Type of ScalarImage is always torchio.INTENSITY')
579			kwargs.update({'type': INTENSITY})
580			super().__init__(args, *kwargs)
581
582
583			class LabelMap(Image):
584			"""Image whose pixel values represent categorical labels.
585
586			Intensity transforms are not applied to these images.
587
588			Example:
589			>>> import torch
590			>>> import torchio as tio
591			>>> labels = tio.LabelMap(tensor=torch.rand(1, 128, 128, 68) > 0.5)
592			>>> labels = tio.LabelMap('t1_seg.nii.gz') # loading from a file
593			>>> tpm = tio.LabelMap( # loading from files
594			... 'gray_matter.nii.gz',
595			... 'white_matter.nii.gz',
596			... 'csf.nii.gz',
597			... )
598
599			See :class:`~torchio.Image` for more information.
600			"""
601			def __init__(self, args, *kwargs):
602			if 'type' in kwargs and kwargs['type'] != LABEL:
603			raise ValueError('Type of LabelMap is always torchio.LABEL')
604			kwargs.update({'type': LABEL})
605			super().__init__(args, *kwargs)
606

fepegar / torchio

Push — master ( e59e2c...7d4e6f )

torchio.data.image.Image.__init__() C

Complexity

Size

Duplication

Importance

How to fix Many Parameters

Many Parameters

Duplication Side-by-Side

Filter issues like

torchio.data.image.Image.init() C