pyalfe.image_processing

class pyalfe.image_processing.ImageProcessor

Bases: ABC

Abstract class for ImageProcessors.

abstract static threshold(image, output, lower_bound, upper_bound, inside_target, outside_target)

Maps any pixel with value inside [lower_bound , upper_bound] to inside_target and any other pixel is mapped to outsize_target. The thresholded image is written to output path.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

• lower_bound (int or float) – The lower_bound for the threshold.

• upper_bound (int or float) – The upper_bound for the threshold.

• inside_target (int or float) –

  The target value for pixels that are inside

  [lower_bound , upper_bound].

• outside_target (int or float) –

  The target value for the pixels that are not inside

  [lower_bound , upper_bound].

static binarize(image, output)

Converts an image to binary by mapping all non-zero values to 1.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

abstract static mask(image, mask, output)

Masks the input image by setting the pixels outside the mask to zero.

Parameters:

• image (str or Path) – Path to the input image.

• mask (str or Path) – Path to the mask.

• output (str or Path) – Path to the output image.

abstract static largest_mask_comp(image, output)

Finds the largest connected component of an input mask image.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

abstract static holefill(binary_image, output)

Fills the holes in a binary image.

Parameters:

• binary_image (str or Path) – Path to the input binary image.

• output (str or Path) – Path to the output image.

abstract static reslice_to_ref(ref_image, moving_image, output)

Reslice the moving_image to the ref_image space.

Parameters:

• ref_image (str or Path) – Path to the reference image.

• moving_image (str or Path) – Path to the moving image.

• output (str or Path) – Path to the output image.

abstract static resample_new_dim(image, output, dim1, dim2, dim3, percent=True)

Resample the image to new dimensions keeping the bounding box

the same, but changing the number of voxels in the image.

Parameters:

• image (str or Path) – Path to the reference image.

• output (str or Path) – Path to the output image.

• dim1 (int) – By default the percentage of the original number of voxels along the first dimension. For example, to double the number of voxel along the first dimension, set dim1 = 200. If percent is False, then dim1 is interpreted as the number of voxels in the output image.

• dim2 (int) – Similar to dim1 but for the second axis.

• dim3 (int) – Similar to dim1 but for the third axis.

• percent (bool) – If True, the dim1, dim2, and dim3 will be interpreted as percentages of the original dimensions instead of raw dimensions.

abstract static get_dims(image)

Parameters:

image (str or Path) – The path to the image.

Returns:

Returns a list containing the dimensions of the image.

Return type:

list

abstract static trim_largest_comp(image, output, trim_margin_vec)

Finds the largest component of the image and trims the margins. This function is usefull for trimming neck.

Parameters:

• image (str or Path) – The path to the image.

• output (str or Path) – Path to the output image.

• trim_margin_vec (tuple or list) – The trim margins along the three dimensions in voxels. For example, (5, 5, 5) means 5 voxels of background are kept in each side of each dimension.

abstract static set_subtract(binary_image_1, binary_image_2, output)

Performs set subtraction between of second binary image from the first binary image.

Parameters:

• binary_image_1 (str or Path) – The path to the first image.

• binary_image_2 (str or Path) – The path to the second image.

• output (str or Path) – Path to the output image.

abstract static dilate(binary_image, rad, output)

Dilates binary image by a certain radius given in voxels.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• rad (int) – The dilation radius in voxels.

• output

abstract static union(binary_image_1, binary_image_2, output)

Takes the union of two binary images.

Parameters:

• binary_image_1 (str or Path) – The path to the first image.

• binary_image_2 (str or Path) – The path to the second image.

• output (str or Path) – Path to the output image.

abstract static distance_transform(binary_image, output)

Computes the distance transform of a binary image.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• output (str or Path) – Path to the output image.

abstract static label_mask_comp(binary_image, output)

Assigns discrete labels to the connected component of a 1 region in a binary image. The largest region is assigned label 1, the second largest is assigned label 2 and so forth.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• output (str or Path) – Path to the output image.

abstract static remap_labels(multi_label_image, label_map, output)

Maps all the labels if a multi label image according to a label map the label_map is a dictionary that maps input labels to output label. For example, {1: 1, 2: 1, 3: 2} maps labels 1 and 2 in the input to label 1 in the output and label 3 in the input to label 2 in the output. Any other label is mapped to 0.

Parameters:

• multi_label_image (str or Path) – The path to the multi label image

• label_map (dict) – The label map

• output (str or Path) – Path to the output image.

class pyalfe.image_processing.Convert3DProcessor

Bases: ImageProcessor

static threshold(image, output, lower_bound, upper_bound, inside_target, outside_target)

Maps any pixel with value inside [lower_bound , upper_bound] to inside_target and any other pixel is mapped to outsize_target. The thresholded image is written to output path.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

• lower_bound (int or float) – The lower_bound for the threshold.

• upper_bound (int or float) – The upper_bound for the threshold.

• inside_target (int or float) –

  The target value for pixels that are inside

  [lower_bound , upper_bound].

• outside_target (int or float) –

  The target value for the pixels that are not inside

  [lower_bound , upper_bound].

static binarize(image, output)

Converts an image to binary by mapping all non-zero values to 1.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

static mask(image, mask, output)

Masks the input image by setting the pixels outside the mask to zero.

Parameters:

• image (str or Path) – Path to the input image.

• mask (str or Path) – Path to the mask.

• output (str or Path) – Path to the output image.

static largest_mask_comp(image, output)

Finds the largest connected component of an input mask image.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

static holefill(binary_image, output)

Fills the holes in a binary image.

Parameters:

• binary_image (str or Path) – Path to the input binary image.

• output (str or Path) – Path to the output image.

static reslice_to_ref(ref_image, moving_image, output)

Reslice the moving_image to the ref_image space.

Parameters:

• ref_image (str or Path) – Path to the reference image.

• moving_image (str or Path) – Path to the moving image.

• output (str or Path) – Path to the output image.

static resample_new_dim(image, output, dim1, dim2, dim3, percent=True)

Resample the image to new dimensions keeping the bounding box

the same, but changing the number of voxels in the image.

Parameters:

• image (str or Path) – Path to the reference image.

• output (str or Path) – Path to the output image.

• dim1 (int) – By default the percentage of the original number of voxels along the first dimension. For example, to double the number of voxel along the first dimension, set dim1 = 200. If percent is False, then dim1 is interpreted as the number of voxels in the output image.

• dim2 (int) – Similar to dim1 but for the second axis.

• dim3 (int) – Similar to dim1 but for the third axis.

• percent (bool) – If True, the dim1, dim2, and dim3 will be interpreted as percentages of the original dimensions instead of raw dimensions.

static get_dims(image)

Parameters:

image (str or Path) – The path to the image.

Returns:

Returns a list containing the dimensions of the image.

Return type:

list

static trim_largest_comp(image, output, trim_margin_vec)

Finds the largest component of the image and trims the margins. This function is usefull for trimming neck.

Parameters:

• image (str or Path) – The path to the image.

• output (str or Path) – Path to the output image.

• trim_margin_vec (tuple or list) – The trim margins along the three dimensions in voxels. For example, (5, 5, 5) means 5 voxels of background are kept in each side of each dimension.

static set_subtract(binary_image_1, binary_image_2, output)

Performs set subtraction between of second binary image from the first binary image.

Parameters:

• binary_image_1 (str or Path) – The path to the first image.

• binary_image_2 (str or Path) – The path to the second image.

• output (str or Path) – Path to the output image.

static dilate(binary_image, rad, output)

Dilates binary image by a certain radius given in voxels.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• rad (int) – The dilation radius in voxels.

• output

static union(binary_image_1, binary_image_2, output)

Takes the union of two binary images.

Parameters:

• binary_image_1 (str or Path) – The path to the first image.

• binary_image_2 (str or Path) – The path to the second image.

• output (str or Path) – Path to the output image.

static distance_transform(binary_image, output)

Computes the distance transform of a binary image.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• output (str or Path) – Path to the output image.

static label_mask_comp(binary_image, output)

Assigns discrete labels to the connected component of a 1 region in a binary image. The largest region is assigned label 1, the second largest is assigned label 2 and so forth.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• output (str or Path) – Path to the output image.

static remap_labels(multi_label_image, label_map, output)

Maps all the labels if a multi label image according to a label map the label_map is a dictionary that maps input labels to output label. For example, {1: 1, 2: 1, 3: 2} maps labels 1 and 2 in the input to label 1 in the output and label 3 in the input to label 2 in the output. Any other label is mapped to 0.

Parameters:

• multi_label_image (str or Path) – The path to the multi label image

• label_map (dict) – The label map

• output (str or Path) – Path to the output image.

class pyalfe.image_processing.NilearnProcessor

Bases: ImageProcessor

static save(nib_image, file)

static crop_img(image, rtol=1e-08, copy=True, pad=(0, 0, 0))

static threshold(image, output, lower_bound, upper_bound, inside_target, outside_target)

Maps any pixel with value inside [lower_bound , upper_bound] to inside_target and any other pixel is mapped to outsize_target. The thresholded image is written to output path.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

• lower_bound (int or float) – The lower_bound for the threshold.

• upper_bound (int or float) – The upper_bound for the threshold.

• inside_target (int or float) –

  The target value for pixels that are inside

  [lower_bound , upper_bound].

• outside_target (int or float) –

  The target value for the pixels that are not inside

  [lower_bound , upper_bound].

static binarize(image, output)

Converts an image to binary by mapping all non-zero values to 1.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

static mask(image, mask, output)

Masks the input image by setting the pixels outside the mask to zero.

Parameters:

• image (str or Path) – Path to the input image.

• mask (str or Path) – Path to the mask.

• output (str or Path) – Path to the output image.

static largest_mask_comp(image, output)

Finds the largest connected component of an input mask image.

Parameters:

• image (str or Path) – Path to the input image.

• output (str or Path) – Path to the output image.

static holefill(binary_image, output)

Fills the holes in a binary image.

Parameters:

• binary_image (str or Path) – Path to the input binary image.

• output (str or Path) – Path to the output image.

static reslice_to_ref(ref_image, moving_image, output)

Reslice the moving_image to the ref_image space.

Parameters:

• ref_image (str or Path) – Path to the reference image.

• moving_image (str or Path) – Path to the moving image.

• output (str or Path) – Path to the output image.

static resample_new_dim(image, output, dim1, dim2, dim3, percent=True)

Resample the image to new dimensions keeping the bounding box

the same, but changing the number of voxels in the image.

Parameters:

• image (str or Path) – Path to the reference image.

• output (str or Path) – Path to the output image.

• dim1 (int) – By default the percentage of the original number of voxels along the first dimension. For example, to double the number of voxel along the first dimension, set dim1 = 200. If percent is False, then dim1 is interpreted as the number of voxels in the output image.

• dim2 (int) – Similar to dim1 but for the second axis.

• dim3 (int) – Similar to dim1 but for the third axis.

• percent (bool) – If True, the dim1, dim2, and dim3 will be interpreted as percentages of the original dimensions instead of raw dimensions.

static get_dims(image)

Parameters:

image (str or Path) – The path to the image.

Returns:

Returns a list containing the dimensions of the image.

Return type:

list

static trim_largest_comp(image, output, trim_margin_vec)

Finds the largest component of the image and trims the margins. This function is usefull for trimming neck.

Parameters:

• image (str or Path) – The path to the image.

• output (str or Path) – Path to the output image.

• trim_margin_vec (tuple or list) – The trim margins along the three dimensions in voxels. For example, (5, 5, 5) means 5 voxels of background are kept in each side of each dimension.

static set_subtract(binary_image_1, binary_image_2, output)

Performs set subtraction between of second binary image from the first binary image.

Parameters:

• binary_image_1 (str or Path) – The path to the first image.

• binary_image_2 (str or Path) – The path to the second image.

• output (str or Path) – Path to the output image.

static dilate(binary_image, rad, output)

Dilates binary image by a certain radius given in voxels.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• rad (int) – The dilation radius in voxels.

• output

static union(binary_image_1, binary_image_2, output)

Takes the union of two binary images.

Parameters:

• binary_image_1 (str or Path) – The path to the first image.

• binary_image_2 (str or Path) – The path to the second image.

• output (str or Path) – Path to the output image.

static distance_transform(binary_image, output)

Computes the distance transform of a binary image.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• output (str or Path) – Path to the output image.

static label_mask_comp(binary_image, output)

Assigns discrete labels to the connected component of a 1 region in a binary image. The largest region is assigned label 1, the second largest is assigned label 2 and so forth.

Parameters:

• binary_image (str or Path) – The path to the binary image.

• output (str or Path) – Path to the output image.

static remap_labels(multi_label_image, label_map, output)

Maps all the labels if a multi label image according to a label map the label_map is a dictionary that maps input labels to output label. For example, {1: 1, 2: 1, 3: 2} maps labels 1 and 2 in the input to label 1 in the output and label 3 in the input to label 2 in the output. Any other label is mapped to 0.

Parameters:

• multi_label_image (str or Path) – The path to the multi label image

• label_map (dict) – The label map

• output (str or Path) – Path to the output image.