multiScaleAnalysis.Tools package

Submodules

multiScaleAnalysis.Tools.PSF_measurements module

Script to analyze the PSF values over an entire image.

multiScaleAnalysis.Tools.PSF_measurements.Gauss(x, a, b, c, d)[source]

Definition of a Gaussian function for fitting: # y = a + (b-a) * np.exp(-(x-c)**2/(2 * d ** 2))

Parameters:
  • x – x variable

  • a – a parameter

  • b – b parameter

  • c – c parameter

  • d – d parameter

multiScaleAnalysis.Tools.PSF_measurements.binarize_label_image(image, threshold)[source]

Binarize and label individual beads in image.

Parameters:
  • image – np.array /3D image to analyze

  • threshold – selected threshold to identify beads

Returns:

np.array with labelled beads

multiScaleAnalysis.Tools.PSF_measurements.determinePSFvalues(croppedimage, lateralstepsize=0.117, axialstepsize=0.2)[source]

Determine the point spread function (PSF) value of a single bead in a cropped image.

Parameters:
  • croppedimage – image containing a single bead.

  • lateralstepsize – Physical dimension of voxel in lateral direction.

  • axialstepsize – Physical dimension of voxel in axial direction.

Returns:

PSF value array: [lateralX_FWHM, lateralY_FWHM, axial_FWHM], returns nan if no Gaussian fit to the function could be found.

multiScaleAnalysis.Tools.PSF_measurements.filter_centroidpositions(centroidlist, areaarray, im_shape, axial_dist=5, lateral_dist=10, areacutoff=300)[source]

Filter the centroid positions to remove beads at the edge and beads that are too big (not point sources) or too small.

Parameters:
  • centroidlist – list of centroid positions

  • areaarray

  • im_shape

  • axial_dist

  • lateral_dist

  • areacutoff – largest area/volume acceptable for a bead

multiScaleAnalysis.Tools.PSF_measurements.filter_psfvalues(psf_list, highestlateralvalue=1, highestaxialvalue=2)[source]

Filter the PSF values to remove clustered beads/aggregates etc.

Parameters:
  • psf_list – list of psf values

  • highestlateralvalue – highest expected lateral value from a single bead

  • highestaxialvalue – highest expected axial value from a single bead

Returns:

filtered list of psf values

multiScaleAnalysis.Tools.PSF_measurements.get_centroidpositions(props)[source]

Return the centroids as an array from region props.

Parameters:

props – Region properties construct

Returns:

np.array(n beads, 3) with centroid positions

multiScaleAnalysis.Tools.PSF_measurements.get_psfvalues(image, centroidlist_filtered, axial_dist=5, lateral_dist=10, lateralstepsize=0.117, axialstepsize=0.2)[source]

Iterate over all beads in an image and generate an image crop for each bead. Calls the determinePSFvalues function to calculate the point spread function (PSF) value of the bead in the image crop, and enter the obtained PSF value into a list of PSF values.

Parameters:
  • image – np.array/image with beads

  • centroidlist_filtered – List of centroid positions.

  • axial_dist – An image crop is generated with extend of centroid_z +/- axial_dist to calculate the PSF.

  • lateral_dist – An image crop is generated with extend of centroid_x +/- lateral_dist and centroid_y +/- lateral_dist to calculate the PSF.

  • lateralstepsize – Physical dimension of voxel in lateral direction.

  • axialstepsize – Physical dimension of voxel in axial direction.

Returns:

An array of PSF values

multiScaleAnalysis.Tools.PSF_measurements.load_image(filename)[source]

Load an image from disk.

Parameters:

filename – filepath to image

Returns:

np.array (image stack)

multiScaleAnalysis.Tools.ShannonEntropyAnalysis module

Script to analyze the Shannon entropy over an entire image series.

multiScaleAnalysis.Tools.ShannonEntropyAnalysis.fast_normalized_dct_shannon_entropy(input_image, frequencycutoff)[source]

Calculates the normalized DCT Shannon entropy of an image. Function from https://github.com/TheDeanLab/navigate/tree/develop/src/navigate/model/analysis

Parameters:
  • input_image – np.ndarray

  • frequencycutoff – determines which percentage of all frequencies are contributing to the entropy (rationale: you don’t want to include noise/high frequencies in the image metric)

Returns:

entropy value

multiScaleAnalysis.Tools.ShannonEntropyAnalysis.load_image(filename)[source]

Load an image from disk.

Parameters:

filename – filepath to image

Returns:

np.array (image stack)

Module contents