image_io module

Wrappers for saving and loading DWIs of 3D printed phantoms.

In this module, images are classified as DWIs or derived images. A DWI should be the raw 4D data from a diffusion MRI scan, and have associated information about the diffusion gradients. A derived image should be (usually) 3D data from analysis of a DWI.

Either of the two may have a mask associated with them.

class image_io.DerivedImage(img)

Bases: object

Wrapper class including an image of data derived from a DWI.

Parameters:img (SpatialImage) – The NiBabel image of the derived data.
get_flat_data()

A 1D numpy array with the image data.

get_image()

A 3D numpy array with the image data.

class image_io.DiffusionWeightedImage(img, gtab)

Bases: object

Wrapper class including image and gradient data.

Parameters:
  • img (SpatialImage) – The NiBabel image of the DWI
  • gtab (GradientTable) – The DIPY gradient table associated with the scan
get_flat_data()

A 1D numpy array with the image data.

get_image()

A 3D numpy array with the image data.

class image_io.MaskedDerivedImage(img, mask)

Bases: image_io.DerivedImage

Wraps an image of data derived from a DWI with a mask.

Parameters:
  • img (SpatialImage) – The NiBabel image of the derived data.
  • mask (array_like) – A 3D binary numpy array, where 1s indicate voxels to be included.
get_flat_data()

A 1D numpy array with the masked derived data.

get_image()

A 3D numpy array with the derived data, ignoring the mask.

class image_io.MaskedDiffusionWeightedImage(img, gtab, mask)

Bases: image_io.DiffusionWeightedImage

Wrapper class including an image, mask, and gradient data.

Parameters:
  • img (SpatialImage) – The NiBabel image of the DWI
  • gtab (GradientTable) – The DIPY gradient table associated with the scan
  • mask (array_like) – A binary numpy array, where 1s indicate voxels to be included.
get_flat_data()

A 1D numpy array with only the masked data.

get_image()

A 3D numpy array with the image data, ignoring the mask.

image_io.gen_table(derived_images)

Organize the model outputs for a phantom in a table.

Parameters:derived_images (collection of MaskedDerivedImages) – A collection of derived images, each corresponding to a single model output from the same phantom. Each of these images must have the same mask associated with them.
Returns:A table where each row corresponds to one voxel, and each column corresponds to one derived image.
Return type:array_like
image_io.load_derived_image(image_path, mask_path=None)

Load the data from a derived image.

Parameters:
  • image_path (string) – Path to the nifti derived data volume.
  • mask_path (string, optional) – Path to the nifti mask, if one exists
Returns:

img – The derived data with a mask, if applicable.
Return type:

DerivedImage
image_io.load_dwi(nifti_path, bval_path, bvec_path, mask_path=None, b0_threshold=250)

Load the data needed to process a diffusion-weighted image.

Parameters:
  • nifti_path (string) – Path to the nifti DWI
  • bval_path (string) – Path to the .bval file
  • bvec_path (string) – Path to the .bvec file
  • mask_path (string, optional) – Path to the nifti mask, if one exists
  • b0_threshold – Threshold below which a b-value is considered zero
Returns:

img – The DWI data with a mask, if applicable.
Return type:

DiffusionWeightedImage
image_io.save_image(data, affine, output_path)

Save some data to a nifti file.

Parameters:
  • data (array_like) – The image data to be saved
  • affine – The affine transform to be used
  • output_path (string) – Path to the file to be saved