napari-simpleitk-image-processing

Process images using SimpleITK in napari

Usage¶

Filters, segmentation algorithms and measurements provided by this napari plugin can be found in the Tools menu. You can recognize them with their suffix (n-SimpleITK) in brackets. Furthermore, it can be used from the napari-assistant graphical user interface. Therefore, just click the menu Tools > Utilities > Assistant (na) or run naparia from the command line.

All filters implemented in this napari plugin are also demonstrated in this notebook.

Gaussian blur¶

Applies a Gaussian blur to an image. This might be useful for denoising, e.g. before applying the Threshold-Otsu method.

Median filter¶

Applies a median filter to an image. Compared to the Gaussian blur this method preserves edges in the image better. It also performs slower.

Bilateral filter¶

The bilateral filter allows denoising an image while preserving edges.

Threshold Otsu¶

Binarizes an image using Otsu's method.

Connected Component Labeling¶

Takes a binary image and labels all objects with individual numbers to produce a label image.

Measurements¶

This function allows determining intensity and shape statistics from labeled images. I can be found in the Tools > Measurement tables menu.

Signed Maurer distance map¶

A distance map (more precise: Signed Maurer Distance Map) can be useful for visualizing distances within binary images between black/white borders. Positive values in this image correspond to a white (value=1) pixel's distance to the next black pixel. Black pixel's (value=0) distance to the next white pixel are represented in this map with negative values.

Binary fill holes¶

Fills holes in a binary image.

Touching objects labeling¶

Starting from a binary image, touching objects can be splits into multiple regions, similar to the Watershed segmentation in ImageJ.

Morphological Watershed¶

The morhological watershed allows to segment images showing membranes. Before segmentation, a filter such as the Gaussian blur or a median filter should be used to eliminate noise. It also makes sense to increase the thickness of membranes using a maximum filter. See this notebook for details.

Watershed-Otsu-Labeling¶

This algorithm uses Otsu's thresholding method in combination with Gaussian blur and the Watershed-algorithm approach to label bright objects such as nuclei in an intensity image. The alogrithm has two sigma parameters and a level parameter which allow you to fine-tune where objects should be cut (spot_sigma) and how smooth outlines should be (outline_sigma). The watershed_level parameter determines how deep an intensity valley between two maxima has to be to differentiate the two maxima. This implementation is similar to Voronoi-Otsu-Labeling in clesperanto.

Richardson-Lucy Deconvolution¶

Richardson-Lucy deconvolution allows to restore image quality if the point-spread-function of the optical system used for acquisition is known or can be approximated.

Installation¶

You can install napari-simpleitk-image-processing via using conda and pip. If you have never used conda before, please go through this tutorial first.

conda install -c conda-forge napari
pip install napari-simpleitk-image-processing

Features¶

The user can select categories of features for feature extraction in the user interface. These categories contain the following measurements:

• size:
  • equivalent_ellipsoid_diameter
  • equivalent_spherical_perimeter
  • equivalent_spherical_radius
  • number_of_pixels
  • number_of_pixels_on_border
• intensity:
  • maximum
  • mean
  • median
  • minimum
  • sigma
  • sum
  • variance
• perimeter:
  • perimeter
  • perimeter_on_border
  • perimeter_on_border_ratio
• shape:
  • elongation
  • feret_diameter
  • flatness
  • roundness
• position:
  • centroid
  • bbox
• moments:
  • principal_axes
  • principal_moments

See also¶

There are other napari plugins with similar functionality for processing images and extracting features:

• morphometrics
• PartSeg
• napari-skimage-regionprops
• napari-cupy-image-processing
• napari-pyclesperanto-assistant
• napari-allencell-segmenter
• RedLionfish
• bbii-decon
• napari-segment-blobs-and-things-with-membranes

Furthermore, there are plugins for postprocessing extracted measurements

• napari-feature-classifier
• napari-clusters-plotter

Contributing¶

Contributions are very welcome. There are many useful algorithms available in SimpleITK. If you want another one available here in this napari plugin, don't hesitate to send a pull-request. This repository just holds wrappers for SimpleITK-functions, see this file for how those wrappers can be written.

License¶

Distributed under the terms of the BSD-3 license, "napari-simpleitk-image-processing" is free and open source software

Citation¶

For implementing this napari plugin, the SimpleITK python notebooks were very helpful. Thus, if you find the plugin useful, consider citing the SimpleITK notebooks:

Z. Yaniv, B. C. Lowekamp, H. J. Johnson, R. Beare, "SimpleITK Image-Analysis Notebooks: a Collaborative Environment for Education and Reproducible Research",
J Digit Imaging., 31(3): 290-303, 2018, https://doi.org/10.1007/s10278-017-0037-8.

Issues¶

If you encounter any problems, please file an issue along with a detailed description.