NiBabel

Access a cacophony of neuro-imaging file formats

Table Of Contents

Previous topic

quaternions

Next topic

volumeutils

Reggie -- the one

spatialimages

A simple spatial image class

The image class maintains the association between a 3D (or greater) array, and an affine transform that maps voxel coordinates to some world space. It also has a header - some standard set of meta-data that is specific to the image format, and extra - a dictionary container for any other metadata.

It has attributes:

  • extra

methods:

  • .get_data()
  • .get_affine() (deprecated, use affine property instead)
  • .get_header() (deprecated, use header property instead)
  • .to_filename(fname) - writes data to filename(s) derived from fname, where the derivation may differ between formats.
  • to_file_map() - save image to files with which the image is already associated.
  • .get_shape() (deprecated)

properties:

  • shape
  • affine
  • header
  • dataobj

classmethods:

  • from_filename(fname) - make instance by loading from filename
  • from_file_map(fmap) - make instance from file map
  • instance_to_filename(img, fname) - save img instance to filename fname.

You cannot slice an image, and trying to slice an image generates an informative TypeError.

There are several ways of writing data.

There is the usual way, which is the default:

img.to_filename(fname)

and that is, to take the data encapsulated by the image and cast it to the datatype the header expects, setting any available header scaling into the header to help the data match.

You can load the data into an image from file with:

img.from_filename(fname)

The image stores its associated files in its file_map attribute. In order to just save an image, for which you know there is an associated filename, or other storage, you can do:

img.to_file_map()

You can get the data out again with:

img.get_data()

Less commonly, for some image types that support it, you might want to fetch out the unscaled array via the object containing the data:

unscaled_data = img.dataoobj.get_unscaled()

Analyze-type images (including nifti) support this, but others may not (MINC, for example).

Sometimes you might to avoid any loss of precision by making the data type the same as the input:

hdr = img.header
hdr.set_data_dtype(data.dtype)
img.to_filename(fname)

Files interface

The image has an attribute file_map. This is a mapping, that has keys corresponding to the file types that an image needs for storage. For example, the Analyze data format needs an image and a header file type for storage:

>>> import nibabel as nib
>>> data = np.arange(24, dtype='f4').reshape((2,3,4))
>>> img = nib.AnalyzeImage(data, np.eye(4))
>>> sorted(img.file_map)
['header', 'image']

The values of file_map are not in fact files but objects with attributes filename, fileobj and pos.

The reason for this interface, is that the contents of files has to contain enough information so that an existing image instance can save itself back to the files pointed to in file_map. When a file holder holds active file-like objects, then these may be affected by the initial file read; in this case, the contains file-like objects need to carry the position at which a write (with to_files) should place the data. The file_map contents should therefore be such, that this will work:

>>> # write an image to files
>>> from io import BytesIO
>>> file_map = nib.AnalyzeImage.make_file_map()
>>> file_map['image'].fileobj = BytesIO()
>>> file_map['header'].fileobj = BytesIO()
>>> img = nib.AnalyzeImage(data, np.eye(4))
>>> img.file_map = file_map
>>> img.to_file_map()
>>> # read it back again from the written files
>>> img2 = nib.AnalyzeImage.from_file_map(file_map)
>>> np.all(img2.get_data() == data)
True
>>> # write, read it again
>>> img2.to_file_map()
>>> img3 = nib.AnalyzeImage.from_file_map(file_map)
>>> np.all(img3.get_data() == data)
True
Header([data_dtype, shape, zooms]) Template class to implement header protocol
HeaderDataError Class to indicate error in getting or setting header data
HeaderTypeError Class to indicate error in parameters into header functions
ImageDataError
ImageFileError
SpatialImage(dataobj, affine[, header, ...]) Initialize image
supported_np_types(obj) Numpy data types that instance obj supports

HeaderDataError

class nibabel.spatialimages.HeaderDataError

Bases: exceptions.Exception

Class to indicate error in getting or setting header data

__init__()

x.__init__(...) initializes x; see help(type(x)) for signature

HeaderTypeError

class nibabel.spatialimages.HeaderTypeError

Bases: exceptions.Exception

Class to indicate error in parameters into header functions

__init__()

x.__init__(...) initializes x; see help(type(x)) for signature

ImageDataError

class nibabel.spatialimages.ImageDataError

Bases: exceptions.Exception

__init__()

x.__init__(...) initializes x; see help(type(x)) for signature

ImageFileError

class nibabel.spatialimages.ImageFileError

Bases: exceptions.Exception

__init__()

x.__init__(...) initializes x; see help(type(x)) for signature

SpatialImage

class nibabel.spatialimages.SpatialImage(dataobj, affine, header=None, extra=None, file_map=None)

Bases: object

Initialize image

The image is a combination of (array, affine matrix, header), with optional metadata in extra, and filename / file-like objects contained in the file_map mapping.

Parameters:

dataobj : object

Object containg image data. It should be some object that retuns an array from np.asanyarray. It should have a shape attribute or property

affine : None or (4,4) array-like

homogenous affine giving relationship between voxel coordinates and world coordinates. Affine can also be None. In this case, obj.affine also returns None, and the affine as written to disk will depend on the file format.

header : None or mapping or header instance, optional

metadata for this image format

extra : None or mapping, optional

metadata to associate with image that cannot be stored in the metadata of this image type

file_map : mapping, optional

mapping giving file information for this image format

__init__(dataobj, affine, header=None, extra=None, file_map=None)

Initialize image

The image is a combination of (array, affine matrix, header), with optional metadata in extra, and filename / file-like objects contained in the file_map mapping.

Parameters:

dataobj : object

Object containg image data. It should be some object that retuns an array from np.asanyarray. It should have a shape attribute or property

affine : None or (4,4) array-like

homogenous affine giving relationship between voxel coordinates and world coordinates. Affine can also be None. In this case, obj.affine also returns None, and the affine as written to disk will depend on the file format.

header : None or mapping or header instance, optional

metadata for this image format

extra : None or mapping, optional

metadata to associate with image that cannot be stored in the metadata of this image type

file_map : mapping, optional

mapping giving file information for this image format

affine
dataobj
files_types = (('image', None),)
classmethod filespec_to_file_map(klass, filespec)

Make file_map for this class from filename filespec

Class method

Parameters:

filespec : str

Filename that might be for this image file type.

Returns:

file_map : dict

file_map dict with (key, value) pairs of (file_type, FileHolder instance), where file_type is a string giving the type of the contained file.

Raises:

ImageFileError :

if filespec is not recognizable as being a filename for this image type.

classmethod filespec_to_files(klass, filespec)
classmethod from_file_map(klass, file_map)
classmethod from_filename(klass, filename)
classmethod from_files(klass, file_map)
classmethod from_filespec(klass, filespec)
classmethod from_image(klass, img)

Class method to create new instance of own class from img

Parameters:

img : spatialimage instance

In fact, an object with the API of spatialimage - specifically dataobj, affine, header and extra.

Returns:

cimg : spatialimage instance

Image, of our own class

get_affine()

Get affine from image

Please use the affine property instead of get_affine; we will deprecate this method in future versions of nibabel.

get_data(caching='fill')

Return image data from image with any necessary scalng applied

The image dataobj property can be an array proxy or an array. An array proxy is an object that knows how to load the image data from disk. An image with an array proxy dataobj is a proxy image; an image with an array in dataobj is an array image.

The default behavior for get_data() on a proxy image is to read the data from the proxy, and store in an internal cache. Future calls to get_data will return the cached array. This is the behavior selected with caching == “fill”`.

Once the data has been cached and returned from an array proxy, if you modify the returned array, you will also modify the cached array (because they are the same array). Regardless of the caching flag, this is always true of an array image.

Parameters:

caching : {‘fill’, ‘unchanged’}, optional

See the Notes section for a detailed explanation. This argument specifies whether the image object should fill in an internal cached reference to the returned image data array. “fill” specifies that the image should fill an internal cached reference if currently empty. Future calls to get_data will return this cached reference. You might prefer “fill” to save the image object from having to reload the array data from disk on each call to get_data. “unchanged” means that the image should not fill in the internal cached reference if the cache is currently empty. You might prefer “unchanged” to “fill” if you want to make sure that the call to get_data does not create an extra (cached) reference to the returned array. In this case it is easier for Python to free the memory from the returned array.

Returns:

data : array

array of image data

See also

uncache
empty the array data cache

Notes

All images have a property dataobj that represents the image array data. Images that have been loaded from files usually do not load the array data from file immediately, in order to reduce image load time and memory use. For these images, dataobj is an array proxy; an object that knows how to load the image array data from file.

By default (caching == “fill”), when you call get_data on a proxy image, we load the array data from disk, store (cache) an internal reference to this array data, and return the array. The next time you call get_data, you will get the cached reference to the array, so we don’t have to load the array data from disk again.

Array images have a dataobj property that already refers to an array in memory, so there is no benefit to caching, and the caching keywords have no effect.

For proxy images, you may not want to fill the cache after reading the data from disk because the cache will hold onto the array memory until the image object is deleted, or you use the image uncache method. If you don’t want to fill the cache, then always use get_data(caching='unchanged'); in this case get_data will not fill the cache (store the reference to the array) if the cache is empty (no reference to the array). If the cache is full, “unchanged” leaves the cache full and returns the cached array reference.

The cache can effect the behavior of the image, because if the cache is full, or you have an array image, then modifying the returned array will modify the result of future calls to get_data(). For example you might do this:

>>> import os
>>> import nibabel as nib
>>> from nibabel.testing import data_path
>>> img_fname = os.path.join(data_path, 'example4d.nii.gz')
>>> img = nib.load(img_fname) # This is a proxy image
>>> nib.is_proxy(img.dataobj)
True

The array is not yet cached by a call to “get_data”, so: >>> img.in_memory False

After we call get_data using the default caching=’fill’, the cache contains a reference to the returned array ``data`:

>>> data = img.get_data()
>>> img.in_memory
True

We modify an element in the returned data array:

>>> data[0, 0, 0, 0]
0
>>> data[0, 0, 0, 0] = 99
>>> data[0, 0, 0, 0]
99

The next time we call ‘get_data’, the method returns the cached reference to the (modified) array:

>>> data_again = img.get_data()
>>> data_again is data
True
>>> data_again[0, 0, 0, 0]
99

If you had initially used caching == ‘unchanged’ then the returned data array would have been loaded from file, but not cached, and:

>>> img = nib.load(img_fname)  # a proxy image again
>>> data = img.get_data(caching='unchanged')
>>> img.in_memory
False
>>> data[0, 0, 0] = 99
>>> data_again = img.get_data(caching='unchanged')
>>> data_again is data
False
>>> data_again[0, 0, 0, 0]
0
get_data_dtype()
get_filename()

Fetch the image filename

Parameters:

None :

Returns:

fname : None or str

Returns None if there is no filename, or a filename string. If an image may have several filenames assoctiated with it (e.g Analyze .img, .hdr pair) then we return the more characteristic filename (the .img filename in the case of Analyze’)

get_header()

Get header from image

Please use the header property instead of get_header; we will deprecate this method in future versions of nibabel.

get_shape()

Return shape for image

This function deprecated; please use the shape property instead

header
header_class

alias of Header

in_memory

True when array data is in memory

classmethod instance_to_filename(klass, img, filename)

Save img in our own format, to name implied by filename

This is a class method

Parameters:

img : spatialimage instance

In fact, an object with the API of spatialimage - specifically dataobj, affine, header and extra.

filename : str

Filename, implying name to which to save image.

classmethod load(klass, filename)
classmethod make_file_map(klass, mapping=None)

Class method to make files holder for this image type

Parameters:

mapping : None or mapping, optional

mapping with keys corresponding to image file types (such as ‘image’, ‘header’ etc, depending on image class) and values that are filenames or file-like. Default is None

Returns:

file_map : dict

dict with string keys given by first entry in tuples in sequence klass.files_types, and values of type FileHolder, where FileHolder objects have default values, other than those given by mapping

set_data_dtype(dtype)
set_filename(filename)

Sets the files in the object from a given filename

The different image formats may check whether the filename has an extension characteristic of the format, and raise an error if not.

Parameters:

filename : str

If the image format only has one file associated with it, this will be the only filename set into the image .file_map attribute. Otherwise, the image instance will try and guess the other filenames from this given filename.

shape
to_file_map(file_map=None)
to_filename(filename)

Write image to files implied by filename string

Parameters:

filename : str

filename to which to save image. We will parse filename with filespec_to_file_map to work out names for image, header etc.

Returns:

None :

to_files(file_map=None)
to_filespec(filename)
uncache()

Delete any cached read of data from proxied data

Remember there are two types of images:

  • array images where the data img.dataobj is an array
  • proxy images where the data img.dataobj is a proxy object

If you call img.get_data() on a proxy image, the result of reading from the proxy gets cached inside the image object, and this cache is what gets returned from the next call to img.get_data(). If you modify the returned data, as in:

data = img.get_data()
data[:] = 42

then the next call to img.get_data() returns the modified array, whether the image is an array image or a proxy image:

assert np.all(img.get_data() == 42)

When you uncache an array image, this has no effect on the return of img.get_data(), but when you uncache a proxy image, the result of img.get_data() returns to its original value.

update_header()

Harmonize header with image data and affine

>>> data = np.zeros((2,3,4))
>>> affine = np.diag([1.0,2.0,3.0,1.0])
>>> img = SpatialImage(data, affine)
>>> img.shape == (2, 3, 4)
True
>>> img.update_header()
>>> img.header.get_data_shape() == (2, 3, 4)
True
>>> img.header.get_zooms()
(1.0, 2.0, 3.0)

supported_np_types

nibabel.spatialimages.supported_np_types(obj)

Numpy data types that instance obj supports

Parameters:

obj : object

Object implementing get_data_dtype and set_data_dtype. The object should raise HeaderDataError for setting unsupported dtypes. The object will likely be a header or a SpatialImage

Returns:

np_types : set

set of numpy types that obj supports