Aurelien Maillot

napari-bud-cell-segmenter






A plugin to segment embryonic mammary bud cells and detect 2 RNA probes

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-bud-cell-segmenter via pip:
pip install napari-bud-cell-segmenter

To install latest development version :
pip install git+https://github.com/AurelienMaillot/napari-bud-cell-segmenter.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-bud-cell-segmenter" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-bud-cell-segmenter

A plugin to segment embryonic mammary bud cells and detect 2 RNA probes













cellfinder
cellfinder is software for automated 3D cell detection in very large 3D images (e.g., serial two-photon or lightsheet volumes of whole mouse brains).
There are three different ways to interact and use it, each with different user interfaces and objectives in mind.
For more details, head over to the documentation website.
At a glance:

There is a command-line interface called brainmapper that integrates with brainreg for automated cell detection and classification. You can install it through brainglobe-workflows.
There is a napari plugin for interacting graphically with the cellfinder tool.
There is a Python API to allow users to integrate BrainGlobe tools into their custom workflows.

Installation
You can find the installation instructions on the BrainGlobe website, which will go into more detail about the installation process if you want to minimise your installation to suit your needs.
However, we recommend that users install cellfinder either through installing BrainGlobe version 1, or (if you also want the command-line interface) installing brainglobe-workflows.
```bash
If you want to install all BrainGlobe tools, including cellfinder, in a consistent manner with one command:
pip install brainglobe>=1.0.0
If you want to install the brainmapper CLI tool as well:
pip install brainglobe-workflows>=1.0.0
```
If you only want the cellfinder package by itself, you can pip install it alone:
bash
pip install cellfinder>=1.0.0
Be sure to specify a version greater than version v1.0.0 - prior to this version the cellfinder package had a very different structure that is incompatible with BrainGlobe version 1 and the other tools in the BrainGlobe suite.
See our blog posts for more information on the release of BrainGlobe version 1.
Seeking help or contributing
We are always happy to help users of our tools, and welcome any contributions. If you would like to get in contact with us for any reason, please see the contact page of our website.
Citation
If you find this package useful, and use it in your research, please cite the following paper:

Tyson, A. L., Rousseau, C. V., Niedworok, C. J., Keshavarzi, S., Tsitoura, C., Cossell, L., Strom, M. and Margrie, T. W. (2021) “A deep learning algorithm for 3D cell detection in whole mouse brain image datasets’ PLOS Computational Biology, 17(5), e1009074
https://doi.org/10.1371/journal.pcbi.1009074

If you use this, or any other tools in the brainglobe suite, please
 let us know, and
 we'd be happy to promote your paper/talk etc.

cellfinder

Automated 3D cell detection in large microscopy images

Ashley Anderson

napari-nibabel






Read access to some common neuroimaging file formats, thanks to the
NiBabel and pydicom
libraries.
Also check out napari-medical-image-formats!

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.
Installation
You can install napari-nibabel via pip:
pip install napari-nibabel

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-nibabel" is free and open source software
Issues
If you encounter any problems, please [file an issue] along with a detailed description.

napari-nibabel

Read access to some common neuroimaging file formats

Lorenzo Gaifas

napari-molecule-reader






A napari plugin that read molecular structure files. It reads PDB and MMCIF files using atomium, expanding molecular assemblies to a full visualization. Data is loaded into napari as Points for ball representation and Vectors for stick representation. If multiple models or assemblies are detected, they will be loaded as separate objects.
https://user-images.githubusercontent.com/23482191/150109390-bd7fb3b4-79b4-43da-aafc-20921714df25.mp4
TODO list:
- [] handle alternate locations (i.e: different conformations in the same pdb model)

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-molecule-reader via pip:
pip install napari-molecule-reader

To install latest development version :
pip install git+https://github.com/brisvag/napari-molecule-reader.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-molecule-reader" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-molecule-reader

A napari plugin that read molecular structure files.

napari-psfgenerator




PSF Generator Napari Plugin
The PSF Generator Napari Plugin provides an intuitive, interactive platform for simulating Point Spread Functions (PSFs) directly within the Napari ecosystem. Built on PyTorch, this plugin supports both CPU and GPU-accelerated computations, ensuring fast and efficient simulations for fundamental and advanced optical modeling.
Key Features:

Flexible Propagation Models: Scalar and vectorial propagators in Cartesian and spherical coordinates.
Customizable Parameters: Configure physical (e.g., numerical aperture, wavelength), numerical (e.g., pixel size, Z-stacks), and optical settings (e.g., Gibson-Lanni corrections, Zernike aberrations).
Real-Time Visualization: Seamless integration with Napari for immediate visual feedback.
Versatile API: Access propagators programmatically for custom workflows.
Image Export: Save computed PSFs in TIFF format.

This plugin is a powerful tool for researchers in optics, computational microscopy, and imaging science, bridging user-friendly interactivity with the computational capabilities of our Python library.

This napari plugin was generated with copier using the napari-plugin-template.

Installation
Set up a Python virtual environment and install napari following this guide.
You can install napari-psfgenerator via pip:
pip install napari-psfgenerator

To install latest development version :
pip install git+https://github.com/Biomedical-Imaging-Group/napari-psfgenerator.git

Now you can try the plugin out! Open napari, click on the menu "Plugins" and select "Propagators (PSF Generator)".
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-psfgenerator" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-psfgenerator

Plugin to compute the focal electric field

Léo Guignard

napari-obj

A plugin to read .obj files

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-obj via pip:
pip install napari-obj

To install latest development version :
pip install git+https://github.com/guignardlab/napari-obj.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-obj" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-obj

Emmanuel Bouilhol

napari-DeepSpot






RNA spot enhancement for fluorescent microscopy images.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-DeepSpot via pip:
pip install napari-DeepSpot

Build from source
This plugin is using Tensorflow, make sure your Python environment has Tensorflow, on create a new environment using the following commands:
* Conda:
conda env create -f environment.yml
conda activate deepspot-napari
* Or pip: 
pip install -r requirements.txt
Usage
Open one or multiple images using Napari GUI : 
File > Open > Select your image
The images are then displayed on Napari
Load the Plugin:
Plugins > Napari-DeepSpot:Enhance Spot

Click on the right panel Button "Enhance"
Wait a few seconds for the magic to happen :

You can see the original images and the enhanced version in the left panel in the layer section.
To save the images : File > Save all layers or File > Save selected layers.

Citation
If you use this plugin please cite the paper:

@article {Bouilhol2021DeepSpot,
   author = {Bouilhol, Emmanuel and Lefevre, Edgar and Dartigues, Benjamin and Brackin, Robyn and Savulescu, Anca Flavia and Nikolski, Macha},
   title = {DeepSpot: a deep neural network for RNA spot enhancement in smFISH microscopy images},
   elocation-id = {2021.11.25.469984},
   year = {2021},
   doi = {10.1101/2021.11.25.469984},
   publisher = {Cold Spring Harbor Laboratory},
   URL = {https://www.biorxiv.org/content/early/2021/11/25/2021.11.25.469984},
   eprint = {https://www.biorxiv.org/content/early/2021/11/25/2021.11.25.469984.full.pdf},
   journal = {bioRxiv}
}  

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-DeepSpot" is free and open source software
Known Issues
If you have troubles with the Python packages typing extensions, use the command :
pip install typing-extensions --upgrade 
When using "Enhance" on multiple images, Napari may freeze. Just wait until it comes to life again, the images will still be enhanced. This is due to Napari memory usage and will be fix one day.
Other Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-DeepSpot

RNA spot enhancement for fluorescent microscopy images

Andrea Bassi

Mark Neil

napari-sim-processor






A Napari plugin for the reconstruction of Structured Illumination Microscopy (SIM) with GPU acceleration (pytorch/cupy if installed).
Currently supports:  
   - conventional SIM data with a generic number of angles and phases (typically, 3 angles and 3 phases are used for resolution improvement in 2D, but any combination can be processed by the widget)
   - hexagonal SIM data with 7 phases, as used in this publication.
   - 3D SIM, for resolution enhancement in three dimensions. This is available in the 3dSIM branch  
The SIM processing widget accepts image stacks organized in 5D (angle,phase,z,y,x).
The reshape widget can be used to easily reshape the data if they are not organized as 5D (angle,phase,z,y,x).
For 3D stacks (raw images) with multiple z-frames, a batch reconstruction method is available, as described here.
Syntetic raw-image stacks of Structured Illumination Microscopy can be easily simulated using the napari SIMulator pluging.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-sim-processor via pip:
pip install napari-sim-processor

To install latest development version :
pip install git+https://github.com/andreabassi78/napari-sim-processor.git

Usage
1) Open napari. 
2) Launch the reshape and sim-processor widgets.
3) Open your raw image stack (using the napari built-in or your own file opener).

4) If your image is ordered as a 5D stack (angle, phase, z-frame, y, x) go to point 6. 
5) In the reshape widget, select the actual number of acquired angles, phases, and frames (red arrow) and press Reshape Stack.
 Note that the label axis of the viewer will be updated (green arrow).

6) In the sim-reconstruction widget press the Select image layer button. Note that the number of phases and angles will be updated (blue arrow). 
7) Choose the correct parameters of the SIM acquisition (NA, pixelsize, M, etc.) and processing parameters (alpha, beta, w, eta, group):
   - w: parameter of the Weiner filter.
   - eta: constant used for calibration. It should be slightly smaller than the carrier frequency (in pupil radius units).
   - group: for stacks with multiple z-frames, it is the number of frames that are used together for the calibration process.
For details on the other parameters see here.
8) Calibrate the SIM processor, pressing the Calibrate button. This will find the carrier frequencies (red circles if the Show Carrier checkbox is selected), the modulation amplitude and the phase, using cross correlation analysis.
9) Click on the checkboxes to show the power spectrum of the raw image (Show power spectrum) or the cross-correlation (Show Xcorr), to see if the found carrier frequency is correct.

Napari viewer showing the power spectrum of the raw stack. The pupil circle is in blue. A circle corresponding to eta is shown in green.

Napari viewer showing the cross-correlation of the raw stack. The red circles indicate the found carrier frequencies
10) Run the reconstruction of a single plane (SIM reconstruction) or of a stack (Stack reconstruction). After execution, a new image_layer will be added to the napari viewer. Click on the Batch reconstruction checkbox in order to process an entire stack in one shot. Click on the pytorch checkbox for gpu acceleration.

Napari viewer with widgets showing a pseudo-widefield reconstruction

Napari viewer with widgets showing a SIM reconstruction
GPU processing
The underlying processing classes will use numpy (and FFTW if available) for 
its calculations. For GPU accelerated processing you need to have either the 
PyTorch (tested with torch v1.11.0+cu113) or the CuPy (tested with cupy-cuda113 
v10.4.0) package installed.  Make sure to match the package cuda version to the CUDA library 
installed on your system otherwise PyTorch will default to CPU and CuPy will not work at all.  
Both packages give significant speedup on even relatively modest CUDA GPUs compared 
to Numpy, and PyTorch running on the CPU only can show improvements relative to numpy 
and FFTW. Selection of which processing package to use is via a ComboBox in the 
napari_sim_processor widget.  Only available packages are shown. 
Other than requiring a CUDA GPU it is advisable to have significant GPU memory 
available, particularly when processing large datasets.  Batch processing is the 
most memory hungry of the methods, but can process 280x512x512 datasets on a 4GB GPU.
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-sim-processor" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-sim-processor

A plugin to process Structured Illumination Microscopy data with gpu acceleration

Hanjin Liu

napari-filaments






A napari plugin for filament analysis.
This plugin helps you to manually track filaments using path shapes of Shapes layer.

Currently implemented functions

Fit paths to filaments in an image as a 2-D spline curve.
Clip/extend paths.
Measurement of filament length at sub-pixel precision.
Basic quantification (mean, std, etc.) along paths.
Import paths from ImageJ ROI file.

Basic Usage
Click Layers > open image to open a tif file. You'll find the image you chose and a shapes layer are added to the layer list.


The "target filaments" box shows the working shapes layer.
The "target image" box shows the image layer on which fitting and quantification will be conducted.
The "filament" box shows currently selected shape (initially this box is empty).

Add path shapes and push key F1 to fit the shape to the filament in the target image layer.

In the "Spline" tab, you can cut/extend or re-fit splines.
In the "Measure" tab, click "measure property" to measure mean instensity etc along each spline.

How it works
Spline fitting is done as following.


This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.
Installation
You can install napari-filaments via pip:
pip install napari-filaments

To install latest development version :
pip install git+https://github.com/hanjinliu/napari-filaments.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-filaments" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-filaments

Titouan Poquillon

napari-yolo5-mitosis-detector






A simple plugin to use yolo5 for mitosis detection with napari
tpoquillon@gmail.com

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-yolo5-mitosis-detector via pip:
pip install napari-yolo5-mitosis-detector

To install latest development version :
pip install git+https://github.com/TPoquillon/napari-yolo5-mitosis-detector.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-yolo5-mitosis-detector" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-yolo5-mitosis-detector

A simple plugin to use yolo5 for mitosis detection with napari

Robert Haase

napari-plugin-search





Find napari plugins

Usage
Enter the name of the plugin you are searching for and use the up and down arrow keys to navigate between them. 
Hit Enter to start a plugin.

This napari plugin was generated with Cookiecutter using with @napari's cookiecutter-napari-plugin template.
Installation
You can install napari-plugin-search via pip:
pip install napari-plugin-search

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-plugin-search" is free and open source software
Issues
If you encounter any problems, please create a thread on image.sc along with a detailed description and tag @haesleinhuepf.

napari-plugin-search

Laura Zigutyte

Ryan Savill

Johannes Müller

Marcelo Zoccoler

Thorsten Wagner

napari-clusters-plotter










A napari-plugin for clustering objects according to their properties.


Note: support for other layers besides the Labels layer as described below is currently only available in the development version. This note will be removed as soon as there is a new release.
Jump to:
- Usage
  - Starting point
  - Measurements
  - Time-Lapse Measurements
  - Plotting
  - Time-Lapse Plotting
  - Dimensionality reduction: UMAP, t-SNE or PCA
  - Clustering
  - Plotting clustering results
- Installation
- Troubleshooting installation
- Contributing
- License
- Acknowledgements
Usage
Starting point
For clustering objects according to their properties, the starting point is a grey-value image and another layer containing derived measurements in the .features property. Here are the supported layer types along with examples of what they may represent:
1. Labels layer containing a label image representing a segmentation of objects.
2. Points layer containing points representing centroid coordinates of objects.
3. Surface layer containing a surface representing a segmentation of an object.
4. Labels layer containing a time-lapse label image representing tracking results, where each label number/color correspond to a unique track ID.
Check the examples data folder to learn how to load these data with code.
1. Labels layer with Segmentation Results
The label image should not contain objects with the label 0 as these objects cannot be separated from the background, which is 0 as well in many images and would lead into erroneous behaviour when performing the clustering.
For segmenting objects, you can for example use the Voronoi-Otsu-labelling approach
in the napari plugin napari-segment-blobs-and-things-with-membranes.

In case you have 2D time-lapse data you need to convert it into a suitable shape using the menu Tools > Utilities > Convert 3D stack to 2D time-lapse (time-slicer) (documentation).
Measurements
The first step is deriving measurements from the labeled image and the corresponding pixels in the grey-value image.
Use the menu Tools > Measurement tables > Regionprops (scikit-image, nsr) to get to the measurement widget (documentation).
Select the image, the corresponding label image and the measurements to analyse and click on Run.
A table with the measurements will open and afterwards, you can save and/or close the measurement table.
At this point it is recommended to close the table and the Measure widget to free space for following steps.
You can also load your own measurements. You can do this using the menu Tools > Measurement > Load from CSV (nsr).
If you load custom measurements, please make sure that there is a label column that specifies which measurement belongs to which labeled object.
Make sure to avoid the label 0 as this is reserved for the background. Tables for time-lapse data need to include an additional column named frame.
Plotting
Once measurements were saved in the labels layer which was analysed, you can then plot these measurements using the menu Tools > Visualization > Plot measurements (ncp).
In this widget, you can select the labels layer which was analysed and the measurements which should be plotted
on the X- and Y-axis. Click on Plot to draw the data points in the plot area.

Under advanced options, you can also select the plot type histogram which will visualize a 2D histogram. 2D histogram visualization is recommended if you have a very high number of data points.

If you choose the same measurement for the X and the Y axis, a histogram will be shown.

Under advanced options you will also find the checkbox determining whether not-selected data points should be hidden (shown in grey) or automatically
clustered as another cluster.
Manual clustering
You can manually select a region in the plot. To use lasso (freehand) tool use left mouse click, and to use a
rectangle - right click. The resulting manual clustering will also be visualized in the original image. To optimize
visualization in the image, turn off the visibility of the analysed labels layer.

Hold down the SHIFT key while annotating regions in the plot to manually select multiple clusters.

Saving manual clustering
Manual clustering results can be saved by going to Tools > Measurement > Show table (nsr), and clicking on Save as csv.
The saved table will contain a "MANUAL_CLUSTER_ID" column. This column is overwritten every time different clusters are manually selected.
Time-Lapse analysis
When you plot your time-lapse datasets you will notice that the plots look slightly different.
Datapoints of the current time frame are highlighted in bright color and you can see the datapoints move through the plot while you navigate through time:

You can also manually select groups using the lasso tool and plot a measurement per frame and see how the group behaves in time.
Furthermore, you could also select a group in time and see where the datapoints lie in a different feature space:

If you have custom measurements from tracking data where each column specifies measurements for a track instead of a label at a specific time point, the frame column must not be added. Check below how tracking data and features should look like.
2. Points Layer with Coordinates
The Points layer typically contains coordinates of objects of interest (for example, object centroids).
To get these coordinates, you can apply spot detection algorithms (check references here with scikit-image, here with pyclesperanto and here for the spotflow plugin) or, if you have segmentation results, use objects centroids as coordinates. This last approach can be done via Tools > Points > Create points from labels centroids (nppas) from the napari-process-points-and-surfaces plugin:

You can load object features to these points by assigning them to the .feature attribute of the Points layer, like this in Python:
points_layer.features = features_table
The number of rows in the table should match the number of points.
You can cluster these features using the same algorithms explained furhter down, or manually, and get points colored accordingly, like shown below:

Check also this notebook to learn how to load these data from code.
3. Surface Layer with Segmentation Results
The Surface layer could contain a surface representing a segmentation result.
To generate this surface from a labeled image containing the segmentation results, a classical algorithm is the Marching Cubes. It is available in scikit-image, and you can also apply it via Tools > Surfaces > Create surface from any label (marching cubes, scikit-image, nppas) from the napari-process-points-and-surfaces plugin. Choose which label id number you want to turn into a surface and click on Run:

You will notice that the surface layer will be created.

You can derive quantitative measurements from the vertices of a surface via Tools > Measurement tables > Surface quality table (vedo, nppas) from napari-process-points-and-surfaces plugin:

Surface vertex measurements can be plotted and classified the same way with the plotter (Tools > Visualization > Plot measurements (ncp)):

Check this notebook to learn how to load these data from code.
4. Labels Layer with Tracking Results
The Labels layer can be also used to display tracking results.
These notebooks show you examples of how to load and format tracking features from Mastodon in a way compatible with napari-clusters-plotter.
For example, if you have a time-lapse labeled image where each label number represents a unique track ID, you can load tracking features to this Labels layer and use the plotter to cluster them. In the 'gif' below, the Tracks layer is NOT used for clustering, it is just shown along as a convenience. There is currently no support for the Tracks layer.

Check this notebook to learn how to load these data from code.
Dimensionality reduction
For getting more insights into your data, you can reduce the dimensionality of the measurements, using these algorithms:
* Uniform Manifold Approximation Projection (UMAP)
* t-distributed stochastic neighbor embedding (t-SNE)
* Principal Component Analysis (PCA)
* Non-linear dimensionality reduction through Isometric Mapping (Isomap)
* Multi-dimensional Scaling (MDS)
To apply them to your data use the menu Tools > Measurement post-processing > Dimensionality reduction (ncp).
Select the label image that was analysed and in the list below, select all measurements that should be dimensionality reduced.
By default, all measurements are selected in the box.
You can read more about parameters of both algorithms by hovering over question marks or by clicking on them.
When you are done with the selection, click on Run and after a moment, the table of measurements will re-appear with two additional columns representing the reduced dimensions of the dataset.
These columns are automatically saved in the labels layer and can be further processed by other plugins.

Afterwards, you can again save and/or close the table.
Clustering
If manual clustering, as shown above, is not an option, you can automatically cluster your data, using these implemented algorithms:
* k-means clustering (KMEANS)
* Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN)
* Gaussian Mixture Model (GMM)
* Mean Shift (MS)
* Agglomerative clustering (AC)
Therefore, click the menu Tools > Measurement post-processing > Clustering (ncp),
select the analysed labels layer.
Select the measurements for clustering, e.g. select only the UMAP measurements.
Select the clustering method KMeans and click on Run.
The table of measurements will reappear with an additional column ALGORITHM_NAME_CLUSTERING_ID containing the cluster
ID of each datapoint.

Afterwards, you can save and/or close the table.
Plotting clustering results
Return to the Plotter widget using the menu Tools > Visualization > Plot measurement (ncp).
Select UMAP_0 and UMAP_1 as X- and Y-axis and the ALGORITHM_NAME_CLUSTERING_ID as Clustering, and click on Plot.

Installation
Devbio-napari installation
The easiest way to install this plugin is to install the devbio-napari plugin collection. The napari-clusters-plotter is part of it.
Minimal installation


Get a python environment, e.g. via mini-conda.
  If you never used mamba/conda environments before, please follow the instructions
  in this blog post first.


Create a new environment, for example, like this:


mamba create --name ncp-env python=3.9

Activate the new environment via conda:

mamba activate ncp-env

Install napari, e.g. via conda:

mamba install -c conda-forge napari
Afterwards, you can install napari-clusters-plotter, e.g. via conda:
mamba install -c conda-forge napari-clusters-plotter
Troubleshooting installation

Error: Could not build wheels for hdbscan which use PEP 517 and cannot be installed directly

This can happen if you used pip for the installation. To solve this error, install hdbscan via conda before installing the plugin:
mamba install -c conda-forge hdbscan

ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 96 from C header, got 88 from PyObject

Similar to the above-described error, this error can occur when importing hdbscan through pip or in the wrong order. This can be fixed by installing packages separately through conda and in the following order:
bash
mamba install -c conda-forge napari hdbscan
pip install napari-clusters-plotter
Contributing
Contributions are very welcome. Tests can be run with pytest, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-clusters-plotter" is free and open source software
Acknowledgements
This project was supported by the Deutsche Forschungsgemeinschaft under Germany’s Excellence Strategy – EXC2068 - Cluster of Excellence "Physics of Life" of TU Dresden.
This project has been made possible in part by grant number 2021-240341 (Napari plugin accelerator grant) from the Chan Zuckerberg Initiative DAF, an advised fund of the Silicon Valley Community Foundation.
Issues
If you encounter any problems, please file an issue along
with a detailed description.

napari-clusters-plotter

A plugin to use with napari for clustering objects according to their properties

Alexis Japas

napari-vesicles-segmentation






A simple plugin to detect vesicles in cells images.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-vesicles-segmentation via pip:
pip install napari-vesicles-segmentation

To install latest development version :
pip install git+https://github.com/alexisjapas/napari-vesicles-segmentation.git

Usage

Open napari
Open your data

Launch the vesicles-segmentation plugin
Select the data you want to segment and set the parameters of the segmentation

image: The image to segment vesicles in. The image can be a 2D or 3D temporal stack of images.
minimum vesicles size: The minimum size of the vesicles to detect. Smaller detected vesicles are removed.
membrane erosion: The size of the disk radius used for eroding the cell. This is used to remove the external membrane. This parameter scales when downsizing the image, for more information see 'downsizing ratio' parameter.
closing size: The size of the disk radius used for closing the cell. This is used to fill holes in the cell. This parameter scales when downsizing the image, for more information see 'downsizing ratio' parameter.
clip: If set to zero, no standardization is performed. Otherwise, the standard deviation of the image is set to n_sigma * the standard deviation of the image, the image is standardized and its values are clipped to the range [-1, 1] in order to remove outliers. The higher the value of n_sigma, the less outliers are removed. This operation can lead to a better detection of the cell.
downsampling ratio: The downsampling ratio used for the downsampled image. This is used to speed up the computation. Downsampling the image have impact in reducing the resolution of erosion and closing e.g. for a downsize ratio of 2, setting the erosion size to 3 will result in an erosion size of 6.
display cell detection: If set to True, the cell detection is displayed in the viewer instead of the vesicle detection.


Click on the "Segment" button to start the segmentation. This can take few seconds or minutes depending on the size of the data. The result is added to the viewer as below.


Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-vesicles-segmentation" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-vesicles-segmentation

A simple plugin to detect vesicles in cells images.

Zach Marin

Talley Lambert

napari-math






This package provides a GUI interfrace for simple mathematical operations on image, point and surface layers.

addition
subtraction
multiplication
division
logical and, or, xor
z-projection (mean and sum)

Operations can be peformed on a single layer or between Image layers (functionaly pending for Surface and Point layers), 
for example adding one layer to another.
When performing operations on two images of different sizes, the result will be the size of the smallest
of the two images.


Installation
You can install napari-math via pip:
pip install napari-math

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-math" is free and open source software
Issues
If you encounter any problems, please file an [issue] along with a detailed description.

napari-math

Simple mathematical operations on image, point and surface layers.

Niklas Netter

napari-nd2-folder-viewer






Look through separate nd2 files in one viewer.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-nd2-folder-viewer via pip:
pip install napari-nd2-folder-viewer

To install latest development version :
pip install git+https://github.com/gatoniel/napari-nd2-folder-viewer.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-nd2-folder-viewer" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-nd2-folder-viewer

Look through separate nd2 files in one viewer.

Jordao Bragantini

ortho-view-napari





It displays the lateral view of the current 3D stack. This could be a starting point for a orthorviewer.

This napari plugin was generated with Cookiecutter using with @napari's cookiecutter-napari-plugin template.

Installation
You can install ortho-view-napari via pip:
pip install ortho-view-napari

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"ortho-view-napari" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

ortho-view-napari

It displays the lateral view of the current 3D stack. This could be a starting point for a orthorviewer.

Saahil Sanganeriya

samcell-napari


A napari plugin for cell segmentation using the Segment Anything Model (SAM) foundation model.

Description
SAMCell-napari provides an intuitive interface for segmenting cells in microscopy images using deep learning. It leverages the power of the Segment Anything Model (SAM) adapted specifically for biological cell segmentation, providing accurate results with minimal tuning.
Key Features:

Simple, user-friendly interface within napari
Compatible with SAMCell models in multiple formats (.pt, .bin, .safetensors)
Support for both SAM-ViT-Base and SAM-ViT-Large model architectures
Adjustable segmentation parameters for fine-tuning
Real-time visualization of results
Distance map visualization for analyzing cell proximity
Full integration with napari's layer system
Enhanced sliding window algorithm with advanced blending for seamless segmentation of large images

What's New in v1.0.0:

Support for multiple model file formats (.pt, .bin, .safetensors)
Improved sliding window algorithm with smooth blending between crops
Better handling of small images and edge cases
Enhanced error recovery and logging
Multiple threshold testing capability
Optimized default thresholds for better segmentation results
Support for both SAM-ViT-Base and SAM-ViT-Large model variants

Installation
You can install samcell-napari via [pip]:
bash
pip install samcell-napari
To install latest development version:
bash
pip install git+https://github.com/saahilsanganeriya/samcell-napari.git
Usage


Start napari
   bash
   napari


Load your image in napari


Open the SAMCell plugin:
   Plugins > samcell-napari > SAMCell Segmentation


Provide the path to your SAMCell model file (supports .pt, .bin, or .safetensors formats)


You can download pre-trained models from the official SAMCell release page


Adjust parameters if needed:

Cell peak threshold: Higher values detect fewer cells (default: 0.47)
Cell fill threshold: Lower values create larger cells (default: 0.09)

Crop size: Size of image crops for processing (default: 256)


Click "Run Segmentation"


View the segmentation results in napari as a Labels layer


Requirements

Python 3.8 or higher
napari 0.4.14 or higher
PyTorch 1.9 or higher
transformers 4.20.0 or higher
CUDA-capable GPU recommended for faster processing

Model Compatibility
The plugin is compatible with SAMCell model files in multiple formats:
- PyTorch model files (.pt)
- Binary model files (.bin) - including the standard pytorch_model.bin
- SafeTensors files (.safetensors) - a safer alternative to PyTorch's pickle-based format
The plugin supports models based on:
- SAM-ViT-Base architecture - Primary model type
- SAM-ViT-Large architecture - Fallback if a model doesn't load with base architecture
Pre-trained models can be downloaded from the official SAMCell release page.
Recommended models include:
- SAMCell1.0-Cellpose-cyto: Trained on the Cellpose cytoplasm dataset
- SAMCell1.0-livecell: Trained on the LiveCELL dataset
These models are part of the release assets for the paper "SAMCell: Generalized Label-Free Biological Cell Segmentation with Segment Anything".
How It Works
SAMCell operates using an enhanced sliding window approach to process large images:

The image is divided into overlapping crops with intelligent handling of image boundaries
Each crop is processed through a SAM-based model
A distance map is created, representing cell centers and boundaries
The crops are stitched back together with smooth blending for seamless transitions
The distance map is processed to extract individual cell masks using watershed segmentation
Results are displayed in napari as labels

Technical Details
Model Type Detection
The plugin intelligently determines the appropriate SAM model architecture:
1. First tries to load the model with SAM-ViT-Base architecture
2. If that fails, automatically falls back to SAM-ViT-Large
3. This ensures maximum compatibility with various pre-trained models
Sliding Window Algorithm
The plugin uses an advanced sliding window algorithm that:
- Handles images of any size, including those smaller than the crop size
- Creates appropriate overlaps between crops to ensure no cells are missed
- Uses a cosine-based blending mask to create smooth transitions between crops
- Fills any potential gaps using nearest neighbor interpolation
Multiple Threshold Testing
For researchers who want to optimize segmentation parameters, the plugin includes a batch processing capability to test multiple threshold combinations at once (available via the API).
Contributing
Contributions are very welcome! Please feel free to submit a Pull Request.
License
Distributed under the MIT License. See LICENSE for more information.
Citation
If you use this plugin in your research, please cite:
@article{samcell2023,
  title={SAMCell: Generalized Label-Free Biological Cell Segmentation with Segment Anything},
  author={...},
  journal={...},
  year={2023}
} 

samcell-napari

A napari plugin for cell segmentation with SAMCell 1.0

Allen Institute for Cell Science

napari-allencell-annotator






A plugin that enables image annotation/scoring and writes annotations to a .csv file. 
Plugin provided by the Allen Institute for Cell Science.
The Allen Cell Image Annotator plugin for napari provides an intuitive
graphical user interface to create annotation templates, annotate large 
image sets using these templates, and save image annotations to a csv file. 
The Allen Cell Image Annotator is a Python-based open source toolkit 
developed at the Allen Institute for Cell Science for both blind, unbiased and un-blind 
microscope image annotating. This toolkit supports easy image set selection
from a file finder and creation of annotation templates (text, checkbox, drop-down, spinbox, and point).
With napari's multi-dimensional image viewing capabilities, the plugin seamlessly allows users to
view each image and write annotations into the custom template.
Annotation templates can be written to a json file for sharing or re-using. After annotating,
the annotation template, image file list, and the annotation values 
are conveniently saved to csv file, which can be re-opened for further annotating. 

Supports the following image types:
OME-TIFF
TIFF
CZI 
PNG 
JPEG
OME-ZARR




This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation using Command Line
1. Prerequisites
The plugin requires Conda.
- Installing on Windows  
  - Follow the steps linked above except
  - On step 8, check top the box to add to PATH
  - 
- Installing on Mac  
2. Install the plugin
Click the link corresponding to your OS.
Windows

From the link above, click the three dots on the top menu bar and select download. 
Open a file explorer and go to the Downloads folder. Use Option 1 below. A prompt window should open and start installing. If this fails use Option 2. 
Option 1: Double-click the file install_napari.sh
Option 2: Search the file finder for Anaconda Prompt. Open version 3. Run the following commands one line at a time. 
conda create -n napari_annotator python=3.10 anaconda
conda activate napari_annotator
python -m pip install --upgrade pip
python -m pip install "napari[all]"
python -m pip install napari-allencell-annotator
napari


Still not working? Try using conda forge instead of pip. 
Ex: conda install -c conda-forge napari instead of python -m pip install "napari[all]"



MacOS/Unix

From the link above, download the file. 
Open terminal. 
Run chmod +x ./Downloads/install_napari.command 
If you get a file not found error try adjusting the path to match where install_napari.command was downloaded.
Open finder, navigate to the file, double-click install_napari.command . 
A terminal window should open and start installing. 

3. Launch the Plugin
Once the napari window opens, go to Plugins.
- If napari-allencell-annotator is listed click it to launch. 
- If it is not listed 
- Install/Uninstall Plugins ⇨ check the box next to napari-allencell-annotator ⇨ close ⇨ Plugins ⇨ napari-allencell-annotator .
4. Re-opening the Plugin After Installing

Windows
Search for anaconda navigator in file finder
Click on navigator version 3
Once the navigator opens, click Environments on the left side
Click on the annotator environment and wait for it to load
Press the play button
Type napari in the prompt that opens
Click Plugins ⇨ napari-allencell-annotator
MacOS
Open terminal
Run these commands one line at a time
conda activate napari_annotator
napari


Click Plugins ⇨ napari-allencell-annotator

Installation from Napari Hub
If you have previously installed Napari on your machine, you can follow these steps to install the plugin from Napari Hub.
1. Install the Plugin

Open Napari
Go to Plugins ⇨ Install/Uninstall Plugins...
Find napari-allencell-annotator in Available Plugins
Click Install
Close the window after the installation finishes

2. Launch the Plugin

Click Plugins ⇨ napari-allencell-annotator
You might have to restart Napari for the annotator to appear in the plugin list.
If you still can't see the plugin, go to Install/Uninstall Plugins ⇨ check the box next to napari-allencell-annotator.

Quick Start

Open napari
Start the plugin 
Open napari, go to Plugins ⇨ napari-allencell-annotator.
Create or import annotations and add images to annotate.

For more detailed usage instructions, check out this document 
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-allencell-annotator" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-allencell-annotator

A plugin that enables annotations provided by Allen Institute for Cell Science

Fabio Steffen

Joel Luethi

napari-ome-zarr-navigator 






A plugin to browse OME-Zarr plates by conditions and load images, labels and features from ROIs
Usage
The ImageBrowser recognizes OME-Zarr plates that were loaded via napari-ome-zarr and allows the selection of wells in the multiwell plate.

The ImageBrowser allows to zoom to a given well ("Go to well") & puts a white bounding box around the selected well.

Using prototype condition tables (to be defined better, see sample data provided by the plugin orthe example in the test data below), the ImageBrowser allows for selecting subsets of the well list based on conditions defined in the condition table. The operetta-compose Fractal task package provides a task to create such condition tables.

The ROI loader (formerly available separately as a napari plugin) can be used standalone or integrated with the ImageBrowser. If a well is selected from the ImageBrowser, all the images in that well can be loaded via the ROI loader.
This supports:
- Loading images from different multiplexing acquisitions
- Loading any ROI based on Fractal ROI tables
- Loading label images
- Loading feature measurements (based on Fractal feature tables in the AnnData format)
This approach of loading label images and feature data has been optimized for and tested with the napari feature classifier.

This plugin is meant to work well for OME-Zarr files generated by Fractal.

Test data
Test data is available at https://zenodo.org/records/11262587
Installation
You can install napari-ome-zarr-navigator via pip:
pip install napari-ome-zarr-navigator

To install latest development version :
pip install git+https://github.com/fractal-analytics-platform/napari-ome-zarr-navigator.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-ome-zarr-navigator" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.
This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

napari-ome-zarr-navigator

A plugin to browse OME-Zarr plates by conditions and load images, labels and features from ROIs

jhnnsrs

mikro-napari






mikro napari enables napari on the mikro/arkitekt platform
DEVELOPMENT
Idea
This is a napari plugin, that provides a simple user interface to use napari with mikro you can view and annotate
data on the mikro platform (synchronised between all of your napari instances) and use napari within arkitekt workflows
(can be extended with other plugins)
Install
Simple install this plugin via naparis plugin-manager and enable it. 
Login with your local mikro/arkitekt platform and start using it in workflows
You can also install mikro-napari directly in your enviroment 
bash
pip install mikro-napari napari[pyqt5]

mikro-napari

A napari plugin to interact with and provide functionality for a connected arkitekt server

Bruno Pannunzio

Marcelo L. Zoccoler

Bruno Schuty

Leonel Malacrida

napari-phasors






A simple plugin to do phasor analysis in napari. Based on the phasorpy library.
Jump to Intallation


Usage
napari-phasors is composed of a few widgets that allow reading a few specific FLIM and hyperspectral file formats, perform phasor analysis, and display and export the results of manual phasor selections.
Sample Data
Two sample datasets for FLIM are provided, along with their corresponding calibration images. Additionally, a paramecium image is included as sample data for hyperspectral analysis.

Phasor Analysis
Plot FLIM Data
FLIM phasor data can be plotted as a 2D histogram or scatter plot. The colormap, the number of bins and the scale of the colors can be customized.
Filtering and thresholding can also be applied to process phasor data and the mean intensity image. 

Plot Hyperspectral Data
Hyperspectral phasor data can also be plotted as a 2D histogram or scatter plot and visualized in the full universal circle.

Apparent Lifetime Display
A FLIM image can be colormapped according to the phase or modulation apparent lifetime. A histogram is also created for visualization of the distribution of apparent lifetimes of the FLIM image.

Phasor Calibration
FLIM images can be calibrated using a reference image acquired under the same experimental parameters. This reference image should consist of a homogeneous solution of a fluorophore with a known fluorescence lifetime and the laser frequency used in the experiment. This ensures accuracy and consistency in lifetime measurements.

Phasor Custom Import
Supported file formats (.tif, .ptu, .sdt, .fbd, .lsm, .ome.tif) can be read and transformed to the phasor space. Additional options, such as the harmonics, channels and frames, can be specified depending on the file format to be read.

Phasor Export
The average intensity image and the phasor coordinates can be exported as OME-TIF files that can be read by napari-phasors and PhasorPy. Alternatively, the phasor coordinates, as well as the selections (cursors) can be exported as a CSV file.

Installation
You can install napari-phasors via pip. Follow these steps from a terminal.
We recommend using miniforge whenever possible. Click here to choose the right download option for your OS.
If you do not use miniforge, but rather Anaconda or Miniconda, replace the mamba term whenever you see it below with conda.
Create a conda environment with napari by typing :
mamba create -n napari-phasors-env napari pyqt python=3.10

Activate the environment :
mamba activate napari-phasors-env

Install napari-phasors via pip :
pip install napari-phasors

Alternatively, install latest development version with :
pip install git+https://github.com/napari-phasors/napari-phasors.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-phasors" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-phasors

napari-update-checker





napari update checker to query for new napari versions.
You can read the documentation at napari.org/update-checker.
Overview
The napari-update-checker is a plugin that checks for newer versions of napari available either on conda-forge, or on PyPI, by querying the repository release tags.

napari-update-checker knows how to detect if napari was installed using conda or pip or if it was installed using the application bundle to provide the correct documentation on how to update napari to the latest version.
Widget

License
Distributed under the terms of the BSD-3 license, "napari-update-checker" is free and open source
software.
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-update-checker

Varun Kapoor

VollSeg Napari Plugin





VollSeg is more than just a single segmentation algorithm; it is a meticulously designed modular segmentation tool tailored to diverse model organisms and imaging methods. While a U-Net might suffice for certain image samples, others might benefit from utilizing StarDist, and some could require a blend of both, potentially coupled with denoising or region of interest models. The pivotal decision left to make is how to select the most appropriate VollSeg configuration for your dataset, a question we comprehensively address in our documentation website.
This project provides the napari plugin for VollSeg, a deep learning based 2D and 3D segmentation tool for irregular shaped cells. VollSeg has originally been developed (see papers) for the segmentation of densely packed membrane labelled cells in challenging images with low signal-to-noise ratios. The plugin allows to apply pretrained and custom trained models from within napari.
For detailed demo of the plugin see these videos and a short video about the parameter selection 

vollseg-napari

Irregular cell shape segmentation using VollSeg

Jonas Windhager

napping





Control point mapping and coordinate transformation using napari
Installation
You can install napping via pip:
pip install napping

To install latest development version:
pip install git+https://github.com/BodenmillerGroup/napping.git

Usage
Run napping for control point mapping and coordinate transformation
Authors
Created and maintained by Jonas Windhager jonas.windhager@uzh.ch
Contributing
Contributing
Changelog
Changelog
License
MIT

napping

Control point mapping and coordination transformation using napari

napari-DEXP





A plugin to interface DEXP with napari.

This napari plugin was generated with Cookiecutter using with @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-dexp via pip:
pip install napari-dexp

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-dexp" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-dexp

napari-ndev










A collection of widgets intended to serve any person seeking to process microscopy images from start to finish, with no coding necessary. napari-ndev was designed to address the gap between the napari viewer and batch python scripting.

Accepts diverse image formats, dimensionality, file size, and maintains key metadata.
Allows advanced, arbitrary image processing workflows to be used by novices.
User-friendly sparse annotation and batch training of machine learning classifiers.
Flexible label measurements, parsing of metadata, and summarization for easily readable datasets.
Designed for ease of use, modification, and reproducibility.

Check out the Docs to learn more!
See the poster presented at BINA 2024 for an overview of the plugins in action
Try out the Virtual I2K 2024 Workshop for an interactive tutorial
Installation
napari-ndev is a pure Python package, and can be installed with pip:
bash
pip install napari-ndev
If napari is currently not installed in your environment, you will also need to include a QtPy backend:
bash
pip install napari-ndev[qtpy-backend]
The easiest way to get started with napari-ndev is to install all the optional dependencies (see note below) with:
bash
pip install napari-ndev[all]

Optional Libraries
napari-ndev is most useful when interacting with some other napari plugins (e.g. napari-assistant) and can read additional filetypes. A few extra BSD3 compatible napari-plugins may be installed with pip:
bash
pip install napari-ndev[extras]
napari-ndev can optionally use GPL-3 licensed libraries to enhance its functionality, but are not required. If you choose to install and use these optional dependencies, you must comply with the GPL-3 license terms. The main functional improvement is from some bioio libraries to support extra image formats, including czi and lif files. These libraries can be installed with pip:
bash
pip install napari-ndev[gpl-extras]
In addition, you may need to install specific bioio readers to support your specific image, such as bioio-czi and bioio-lif (included in [gpl-extras]) or bioio-bioformats.
Development Libraries
For development use the [dev] optional libraries to verify your changes, which includes the [docs] and [testing] optional groups. However, the Github-CI will test pull requests with [testing] only.

The wide breadth of this plugin's scope is only made possible by the amazing libraries and plugins from the python and napari community, especially Robert Haase.
This napari plugin was generated with Cookiecutter using napari's cookiecutter-napari-plugin template.
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-ndev" is free and open source software.
Some optional libraries can be installed to add functionality to napari-ndev, including some that may be more restrictive than this package's BSD-3-Clause.
Issues
If you encounter any problems, please [file an issue] along with a detailed description.

napari-ndev

napari widgets to (batch) process images from start to finish.

Herearii Metuarea

napari-aphid






A plugin to classify aphids by stage of development.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-aphid via pip:
pip install napari-aphid

To install latest development version :
pip install git+https://github.com/hereariim/napari-aphid.git

Description
This plugin is a tool to count the number of aphids from two models developed on ilastik. Implemented in napari, this tool allows the correction of pixels and labels that are not well 
predicted. 
In this plugin we find our two main parts of the aphid counting model presented in two widgets. A third widget allows to save the updates applied on the segmentation mask.
This plugin is an use cas, dedicated to private use of french laboratory.
Plugin input
Segmentation
The user must give two objects as input:

Compressed file in .zip format
Ilastik pixel classification model in .ilp format

In particular, compressed file must be organized as follows:
.
└── Country.zip
    └── Country
        ├── Area1
        │   ├── Area1.im_1.tif
        │   ├── Area1.im_1.h5
        │   ├── Area1.im_2.tif 
        │   ├── Area1.im_2.h5  
        │   ├── Area1.im_3.tif
        │   ├── Area1.im_3.h5
        │   ...
        │   ├── Area1.im_n.tif
        │   └── Area1.im_n.h5
        │
        ├── Area2
        │   ├── Area2.im_1.tif
        │   ├── Area2.im_1.h5
        │   ├── Area2.im_2.tif
        │   ├── Area2.im_2.h5
        │   ├── Area2.im_3.tif
        │   ├── Area2.im_3.h5
        │   ...
        │   ├── Area2.im_n.tif
        │   └── Area2.im_n.h5
        │
        ...
        │
        └── Arean
            ├── Arean.im_1.tif
            ├── Arean.im_1.h5
            ├── Arean.im_2.tif
            ├── Arean.im_2.h5
            ├── Arean.im_3.tif
            ├── Arean.im_3.h5
            ...
            ├── Arean.im_n.tif
            └── Arean.im_n.h5
In each folder Area1, Area2, ..., Arean, we notice that each tif image is accompanied by its h5 version. The images in h5 format were generated by the Export h5 widget of the Ilastik plugin in the ImageJ software.
Classification
The user must give the Ilastik object classification model in .ilp format.
Widget: Image segmentation
This widget is a tool to segment a set of images. It takes as input a compressed file of images and an ilastik segmentation model. A Run button is used to start the image segmentation process. In the background, the console presents the progress status. This widget returns a menu which is a list of processed images. This list allows an RGB image and its segmentation mask to be displayed in the napari window.

User conduct : In this widget, the user corrects the image with the annotation tools (brush and eraser only). With the brush, he/she has to add the same colour presented in the image. To obtain this colour, the user can take the color with the color picker tool. With the eraser, he/she erase colour not well predicted. Tous les annotations appliquées dans l'image doit être sauvegarder avec le bouton Save du widget Save modification
Widget: Save modification
This is the backup of the segmentation mask. It saves updates applied to the mask.
Widget: Object classification
This widget is a tool to classify segmented images. It takes as input an ilastik object classification model. A Run button is used to start the classification process. In the background, the console shows the progress of the image processing. This widget returns a menu that lists the processed images. This list provides two elements. The first is the display of the selected image in the window. The second is the display of a table that shows the predicted classes for each object.

User conduct : In this widget, the user corrects labels not well predicted in the table at the bottom right. He must not forget to save his correction with the Save button.
When the user has finished with all his images, he uses the Export button to import a quantitative table. This table contains for each image, the name of the aphid type and its size in pixels.
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-aphid" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-aphid

A plugin to classify aphids by stage of development.

napari-segment-blobs-and-things-with-membranes (nsbatwm)








This napari-plugin is based on scikit-image and allows segmenting nuclei and cells based on fluorescence microscopy images with high intensity in nuclei and/or membranes.
Usage
This plugin populates image processing operations to the Tools menu in napari.
You can recognize them with their suffix (nsbatwm) 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.

You can also call these functions as shown in the demo notebook.
Voronoi-Otsu-Labeling
This algorithm uses Otsu's thresholding method in combination with 
Gaussian blur and a 
Voronoi-Tesselation 
approach to label bright objects such as nuclei in an intensity image. The alogrithm has two sigma parameters which allow
you to fine-tune where objects should be cut (spot_sigma) and how smooth outlines should be (outline_sigma).
This implementation aims to be similar to Voronoi-Otsu-Labeling in clesperanto.

Seeded Watershed
Starting from an image showing high-intensity membranes and a seed-image where objects have been labeled (e.g. using Voronoi-Otsu-Labeling),
objects are labeled that are constrained by the membranes.

Seeded Watershed with mask
If there is additionally a mask image available, one can use the Seeded Watershed with mask, to constraint the flooding 
on a membrane image (1), starting from nuclei (2), limited by a mask image (3) to produce a cell segmentation within the mask (4).

Seeded Watershed using local minima as starting points
Similar to the Seeded Watershed and Voronoi-Otsu-Labeling explained above, you can use this tool to segment an image
showing membranes without an additional image showing nuclei. The two sigma parameters allow to fine tune how close 
objects can be and how precise their boundaries are detected.

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

Subtract background
Subtracts background using scikit-image's rolling-ball algorithm. 
This might be useful, for example to make intensity of membranes more similar in different regions of an image.

Threshold Otsu
Binarizes an image using scikit-image's threshold Otsu algorithm, also known as 
Otsu's method.

Split touching objects (formerly known as binary watershed).
In case objects stick together after thresholding, this tool might help.
It aims to deliver similar results as ImageJ's watershed implementation.

Connected component labeling
Takes a binary image and produces a label image with all separated objects labeled differently. Under the hood, it uses
scikit-image's label function.

Manual split and merge labels
Split and merge labels in napari manually via the Tools > Utilities menu:


This napari plugin was generated with Cookiecutter using with @napari's cookiecutter-napari-plugin template.
Installation
This plugin is part of devbio-napari. To install it, please follow its installation instructions.
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-segment-blobs-and-things-with-membranes" is free and open source software
Issues
If you encounter any problems, please create a thread on image.sc along with a detailed description and tag @haesleinhuepf.

napari-segment-blobs-and-things-with-membranes

A plugin based on scikit-image for segmenting nuclei and cells based on fluorescent microscopy images with high intensity in nuclei and/or membranes

Clément Cazorla

# napari-svetlana








The aim of this plugin is to classify the output of a segmentation algorithm.
The inputs are :

A folder of raw images
Their segmentation masks where each ROI has its own label.

Svetlana can process 2D, 3D and multichannel image. If you want to use it to work on cell images, we strongly
recommend the use of Cellpose for the segmentation part, as it provides excellent quality results and a standard output format
accepted by Svetlana (labels masks). 
If you use this plugin please cite the paper: 
Cazorla, C., Weiss, P., & Morin, R. (2024). Svetlana: a Supervised Segmentation Classifier for Napari.
```bibtex
@article{cazorla2024svetlana,
  title={Svetlana a supervised segmentation classifier for Napari},
  author={Cazorla, Cl{\'e}ment and Morin, Renaud and Weiss, Pierre},
  journal={Scientific Reports},
  volume={14},
  number={1},
  pages={11604},
  year={2024},
  publisher={Nature Publishing Group UK London}
}
```


This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
First install Napari in a Python 3.9 Conda environment following these instructions :
bash
conda create -n svetlana_env python=3.9
conda activate svetlana_env
conda install pip
python -m pip install "napari[all]"==0.4.17
Then, you can install napari_svetlana via pip, or directly from the Napari plugin manager (see Napari documentation):
bash
pip install napari_svetlana
WARNING:
If you have a Cuda compatible GPU on your computer, some computations may be accelerated
using Cupy. Unfortunately, Cupy needs Cudatoolkit to be installed. This library can only be installed via 
Conda while the plugin is a pip plugin, so it must be installed manually for the moment:
bash
conda install cudatoolkit=11.5
Also note that the library (Cucim) that we use to improve these performances, computing morphological operations on GPU
is unfortunately only available for Linux systems. Hence, if you are a Windows user, this installation is not necessary.
Tutorial
Many advanced features are available in Svetlana, such as data augmentation or contextual information reduction, to optimize the performance of your classifier. Thus, we strongly encourage you to
check our Youtube tutorial and
our documentation.
A button called TRY ON DEMO IMAGE is available in the annotation plugin and enables you to apply the YouTube
tutorial to the same test images to learn how to use the plugin. Feel free to try it to test all the features
that Svetlana offers.
Similar Napari plugins
Joel Luethi developed a similar method for objects classification called napari feature classifier.
Also, apoc by Robert Haase is available in Napari for pixels and objects classification.
Contributing
Contributions are very welcome.
License
Distributed under the terms of the BSD-3 license,
"napari_svetlana" is free and open source software
Acknowledgements
The method was developed by Clément Cazorla, Renaud Morin and Pierre Weiss. And the plugin was written by
Clément Cazorla. The project is co-funded by Imactiv-3D and CNRS.
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-svetlana

A classification plugin for the ROIs of a segmentation mask. If you face problems opening the Napari-hub page, try replacing napari-svetlana by napari_svetlana in the URL.

Constantin Aronssohn

Arthur Masson

napari-exodeepfinder






A napari plugin for the ExoDeepFinder library which includes display, annotation, target generation, segmentation and clustering functionalities.
An orthoslice view has been added for an easier visualisation and annotation process.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation

Create a conda environment with python 3.10: conda create -n napari-exodeepfinder python=3.10
Activate the environment: conda activate napari-exodeepfinder
Install napari: pip install napari-exodeepfinder

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the GNU GPL v3.0 license,
"napari-exodeepfinder" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-exodeepfinder

A napari plugin for the ExpDeepFinder library which includes display, annotation, target generation, segmentation and clustering functionalities. An orthoslice view has been added for an easier visualisation and annotation process.

Thomas Bonte

Cut Detector






Automatic micro-tubule cut detector.
https://github.com/user-attachments/assets/2af2e1a6-adf9-4d63-a353-e190c4814d83

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.



  Your browser does not support the video tag.

Installation
Conda environment
It is highly recommended to create a dedicated conda environment, by following these few steps:


Install an Anaconda distribution of Python. Note you might need to use an anaconda prompt if you did not add anaconda to the path.


Open an Anaconda prompt as admin to create a new environment using conda. We advice to use python 3.10 and conda 23.10.0, to get conda-libmamba-solver as default solver.


bash
conda create --name cut_detector python=3.10 conda=23.10.0
conda activate cut_detector
Package installation
Once in a dedicated environment, our package can be installed via pip:
bash
pip install cut_detector
Alternatively, you can clone the github repo to access to playground scripts.
bash
git clone https://github.com/15bonte/cut-detector.git
cd cut-detector
pip install -e .
GPU
We highly recommend to use GPU to speed up segmentation. To use your NVIDIA GPU, the first step is to download the dedicated driver from NVIDIA.
Next we need to remove the CPU version of torch:
bash
pip uninstall torch
The GPU version of torch to be installed can be found here. You may choose the CUDA version supported by your GPU, and install it with conda. This package has been developed with the version 11.6, installed with this command:
bash
conda install pytorch==1.12.1 torchvision pytorch-cuda=11.6 -c pytorch -c nvidia
Update
To update cut-detector to the latest version, open an Anaconda prompt and use the following commands:
bash
conda activate cut_detector
pip install cut-detector --upgrade
Definitions
Each detected cell division is labeled with one of the following categories:

NORMAL: Division happening as expected, where (at least) 1 micro-tubule cut is detected.
NO_MID_BODY_DETECTED: Along the cell division, no mid-body was detected on the MKLP1 channel. This category encompasses different cases: the detection may have failed, the mid-body may not express the fluorescence, or this may not actually be a division.
MORE_THAN_TWO_DAUGHTER_TRACKS: Tripolar division. This category encompasses both actual tripolar divisions and wrong identifications of daughter cells (mainly caused by segmentation issues).
NEAR_BORDER: Division close to the border of the image, hence ignored as it is likely to be difficult to detect micro-tubule cuts. A division is classified as NEAR_BORDER as soon as the distance between 1 detected mid-body and the border of the image is less than 20px.
NO_CUT_DETECTED: Division whose mid-body was detected, but with all micro-tubule bridges classified as "No cut". Likely to be at the end of the video, cells dying before the end of division, or cells going out of frame.
TOO_SHORT_CUT: First micro-tubule cut detected before or at 50 minutes. Ignored as this is very unlikely, so it is probably caused by a wrong division detection.

Division movies start at the maximum between:

Mother cell start frame
10 frames before the end of metaphase

Division movies end at the minimum between:

Last frame of any of the daughter cells
Metaphase of any of the daughter cells

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
Scripts required to improve any of Cut Detector tasks can be found in the folder developers.
License
Distributed under the terms of the BSD-3 license,
"cut-detector" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

cut-detector

Lars Krämer

napari-labels
A plugin for flexible and intuitive color management of label layers in Napari, designed to enhance the visual clarity and control of your labeled data.
Assign custom colors to individual classes using a color picker or RGB input, adjust per-class opacity, and apply colormaps to visually distinguish between labels with ease.
Installation
You can install napari-labels via pip:
pip install napari-labels
Seaborn Color palettes
You can use all color palette names which are valid for seaborn.color_palette().
An overview can be found here:

https://r02b.github.io/seaborn_palettes/
https://www.practicalpythonfordatascience.com/ap_seaborn_palette

License
Distributed under the terms of the Apache Software License 2.0 license,
"napari-labels" is free and open source software
Acknowledgments

     
  

This repository is developed and maintained by the Applied Computer Vision Lab (ACVL)
of Helmholtz Imaging and the
Division of Medical Image Computing at DKFZ.
This napari plugin was generated with copier using the napari-plugin-template.

napari-labels

Adding intuitive label color management, allowing custom class colors, per-class opacity, and colormap control to enhance visual clarity.

napari-curtain






View one image over another as curtain

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.
Usage
You find the Curtain plugin in the menu Tools > Visualization > Curtain. Move the position of the slider left/right 
as shown in the video above. In case one image is much bright than the other, you can modify the factors above the 
slider until visualization pleases.
Installation
You can install napari-curtain via pip:
pip install napari-curtain

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-curtain" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-curtain

lennart kowitz

MSI-Explorer






User Manual
The MSI-Explorer napari plugin is a powerful tool designed for targeted biochemical annotations in MSI data. This user manual provides a comprehensive guide on how to install, use, and explore the functionalities of the plugin within the napari platform. It covers data import, visualization, mean intensity calculation, region of interest (ROI) analysis, annotation with selected databases and pre-processing such as noise reduction and normalization. 
[MSI-Explorer] 
Installation
Install napari by using this command.
 pip install "napari[all]"

You can install MSI-Explorer via pip:
 pip install MSI-Explorer

Usage
Start napari from the console with:
napari

Navigate to Plugins -> MSI-Explorer (MSI-Explorer)

1. Uploading and visualization of mass spectrometry imaging data

Select imzml file using Load imzML.
Metadata can be checked by View Metadata.




Upon uploading profile mode data, a pop-up appears prompting you to convert it to centroid mode.
Selecting Yes converts the data, while No keeps it in its original profile format.

2. Calculating mean (average) intensity

To calculate the mean spectrum, click on Show true mean spectrum.
Clicking Show image will create an image view of the currently plotted data
To export the plotted data as .csv file, click Export spectrum data.
To save the spectrum plot as image, click Export spectrum plot.



2.1. Calculating mean (average) intensity of selected m/z value
To focus on a specific m/z value, zoom in on the spectrum plot. The figure will be as
shown as below.

It is recommended to use Multi panel view.
The image can be displayed by Show image and the data can be exported as .csv file by using Export spectrum data.
3. Pre-processing
The pre-processing capabilities of MSI-Explorer enhance data quality and prepare MSI data for downstream analysis. Pre-processing steps involve: 
(a) Noise reduction
Users can choose their desired level of noise reduction (shown as a percentage) for their experiment. 

(b) Normalization
The normalization methods that the user can apply are 
- Total ion current (TIC)
- Root mean square (RMS)
- Medium
- Reference peak (or internal standard)

(c) Hotspot removal
Hotspot removal can also be applied using a default threshold of 99.99%.

After pre-processing steps are chosen, click Execute and Show true mean spectrum to calculate the mean intensity.
The figure shows the spectrum and image of the TIC normalization with 3% noise reduction and hotspot removal for the 99.9% quantile.

4. Database
To use the database search, click on Select and a pop-up window will appear. There,
select Metabolite_database_ver2, which is a built-in database, and click Confirm.

The features of the database function are
1. Charge (neutral, positive or negative)
2. Adduct (based on the charge chosen)
3. Range of the m/z value for the image display
4. custom search with molecule name or m/z value

Users can customize the database with exact mass, molecule name, or molecular formula. The format should be as shown in the table and the headers are not needed in the database.
Exact mass | Molecule name | Molecula formula
------- | -------- | --------
176.0950 | Cotinine | C10H12N2O
174.1117 | Arginine | C6H14N4O2
244.0881 | Biotin | C10H16N2O3S
5. Region of interest (ROI) selection

To select the ROI, click on Select ROI for mean spectrum. Adjust the brush size and label color. You can fill the area by using paint icon. 
Then click on the Calculate ROI mean spectrum.
You can export as .csv file by using Export spectrum data.
If one m/z is needed, just zoom-in the spectrum plot window and export.
Before selecting the second ROI, remove the first selected area by using eraser or label 0.


Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"MSI-Explorer" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

MSI-Explorer

a napari plug-in for biochemical annotation of mass spectrometry imaging data


blik








blik is a tool for visualising and interacting with cryo-ET and subtomogram averaging data. It leverages the fast, multi-dimensional napari viewer and the scientific python stack.
DISCLAIMER: this package is in development phase. Expect bugs and crashes. Please, report them on the issue tracker and ask if anything is unclear!
Installation
You can either install blik through the napari plugin system, through pip, or get both napari and blik directly with:
bash
pip install "blik[all]"
The [all] qualifier also installs pyqt5 as the napari GUI backend, and a few additional napari plugins that you might find useful in your workflow:
- napari-properties-plotter
- napari-properties-viewer
- napari-label-interpolator
Nightly build
If you'd like the most up to date blik possible, you can install directly from the main branch on github. This also uses napari main, so expect some instability!
pip install "git+https://github.com/brisvag/blik.git@main#egg=blik[all]"
pip install "git+https://github.com/napari/napari.git@main#egg=napari[all]"
Basic Usage
From the command line:
bash
napari -w blik -- /path/to.star /path/to/mrc/files/*
The -w blik is important for proper initialization of all the layers. Always open the main widget open to ensure nothing goes wrong!
blik is just napari. Particles and images are exposed as simple napari layers, which can be analysed and manipulated with simple python, and most importantly other napari plugins.
Widgets
The main widget has a few functions:

experiment: quickly switch to a different experiment id (typically, everything related to an individual tomogram such as volume, particles and segmentations)
new: generate a new segmentation, a new manually-picked set of particles, or a new surface, sphere, or filament picking for segmentation, particle generation or volume resampling.
add to exp: add a layer to the currently selected experiment (just a shorthand for layer.metadata['experiment_id'] = current_exp_id)
slice_thickness: changes the slicing thickness in all dimensions in napari. Images will be averaged over that thickness, and all particles in the slice will be displayed.

There are also widgets for picking surfaces, spheres and filaments:

surface: process a previously picked surface picking layer to generate a surface mesh and distribute particles on it for subtomogram averaging, or resample a tomogram along the surface.
sphere: process a previously picked sphere picking layer to generate a sphere mesh and distribute particles on it for subtomogram averaging.
filament: process a previously picked filament picking layer to generate a filament and distribute particles on it for subtomogram averaging, or resample a tomogram along the filament.

References
If you use blik, please cite the repo on zenodo and the paper on Plos Biology: https://doi.org/10.1371/journal.pbio.3002447.

blik

Python tool for visualising and interacting with cryo-ET and subtomogram averaging data.

Krishnan Venkataraman

napari-SAMV2
Napari plugin to use segment anything version 2.1 models from Meta.
Plugin made for segmenting 3d volumetric data or 3d time series data.

Installation
The plugin requires the following pre-requisite to be installed :

Python and pytorch versions

python>=3.10,torch>=2.5.1 and torchvision>=0.20.1 required
To install pytorch with your respective OS please visit - https://pytorch.org/get-started/locally/

SAM v2 installation from meta

Please refer https://github.com/facebookresearch/sam2

Install napari

python -m pip install "napari[all]"
Following is a sample conda environment installation with the above pre-req 
conda create -n samv2_env python=3.10
conda activate samv2_env
pip3 install torch torchvision

git clone https://github.com/facebookresearch/sam2.git && cd sam2
pip install -e .

python -m pip install "napari[all]"

pip install napari-SAMV2

Usage
Middle mouse click - positive point or keyboard shortcut "a"
Ctrl + Middle mouse click - negative point or keyboard shortcut "n"
Time Series Segmentation :

Volume Segmentation :

Reference :
Example Data in the demo videos are from,
Cell tracking challenge - https://celltrackingchallenge.net/ 
and
FlyEM project - https://www.janelia.org/project-team/flyem/hemibrain
License
Distributed under the terms of the BSD-3 license,
"napari-SAMV2" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-SAMV2

Napari plugin for segment anything version 2 model from meta. Plugin primarily useful for segmenting 3d volumetric data or 3d time series data. 

Chenqi Zhang

napari-aideveloper






napari_aideveloper is a napari-plugin derived from AIDeveloper that allows you to train, evaluate, and apply deep neural nets for image classification within a graphical user-interface (GUI).

Installation
You can install napari-aideveloper via pip:
pip install napari-aideveloper

Introduction
Main functions

Build
History


Build
1. Load data
Drag and drop your data in .rtdc (HDF5) format into the file table and set the class and training/validation.

2. Choose Neural Networks

3. Set model storage path

4. Start fitting


Preview image

Image augmentation


History
1. Load meta data

2. Check model details

3. Rolling median & Linear fit

Contributing
Contributions are very welcome. You can submit your pull request on GitHub. Tests can be run with tox, please ensure the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-aideveloper" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

napari-aideveloper

napari_aideveloper is a napari-plugin deived from AIDeveloper that allows you to train, evaluate and apply deep neural nets for image classification within a graphical user-interface (GUI).

Tim-Oliver Buchholz

napari-psf-analysis






This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.



A plugin to analyse point spread functions (PSFs) of optical systems.

Usage
Starting Point
To run a PSF analysis open an image of acquired beads. Add a point-layer
and indicate the beads you want to measure by adding a point.
Run Analysis
Open the plugin (Plugins > napari-psf-analysis > PSF Analysis) and ensure
that your bead image and point layers are select in the Basic tab under
Image and Points respectively.
In the Advanced tab further information can be provided. Only the filled
in fields of the Advanced tab are saved in the output.
After verifying all input fields click Extract PSFs.
Discard and Save Measurement
Once the PSF extraction has finished a new layer (Analyzed Beads) appears,
holding a summary
image for every selected bead.
Individual summaries can be discarded by clicking the Delete Displayed
Measurement button.
Results are saved to the selected Save Dir by clicking the Save
Measurements button.
Note: Beads for which the bounding box does not fit within the image are
automatically excluded from the analysis and no output is generated.
Saved Data
Every image of the Analyzed Beads layer is saved as {source_image_name}_X
{bead-centroid-x}_Y{bead-centroid-y}_Z{bead-centroid-z}.png file.
Additionally a PSFMeasurement_{source_image_acquisition_date}_
{source_image_name}_{microscope_name}_{magnification}_{NA}.csv file is
stored containing the measured values and all filled in fields.


Installation
We recommend installation into a fresh conda environment.
1. Install napari
```shell
conda create -y -n psf-analysis -c conda-forge python=3.9
conda activate psf-analysis
conda install -c conda-forge napari pyqt
```
2. Install napari-aicsimageio and bioformats
Required if you want to open other files than .tif e.g. .stk..
Note: See napari-aicsimageio for more information about opening images.
```shell
conda install -c conda-forge openjdk bioformats_jar "aicsimageio[all]" napari-aicsimageio
conda deactivate
conda activate psf-analysis
```
3. Install napari-psf-analysis
You can install napari-psf-analysis via pip:
shell
python -m pip install xmlschema
python -m pip install napari-psf-analysis
4. Optional Set Config
You can provide a config yaml file with the available microscope names and a default save directory.
This will change the Microscope text field to a drop down menu and change the default save directory.
example_config.yaml
yaml
microscopes:
  - TIRF
  - Zeiss Z1
output_path: "D:\\psf_analysis\\measurements"
To use this config navigate to Plugins > napari-psf-analysis > Set Config and select the config file.
Note: The save path is OS specific.

5. Desktop Icon for Windows
Follow these instructions by Robert Haase.
Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-psf-analysis" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-psf-analysis

A plugin to analyse point spread functions (PSFs).

napari-animated-gif-io






Open and save 3D image stacks as animated gifs
You find the menus for opening and saving animated gifs in the Tools > File Import/Export menu:

Furthermore, if in 3D view, you can save the current view with a little tilt animation as animated gif.
Under the hood this uses the microfilm library.


This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.
Installation
You can install napari-animated-gif-io via pip:
pip install napari-animated-gif-io

To install latest development version :
pip install git+https://github.com/haesleinhuepf/napari-animated-gif-io.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-animated-gif-io" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-animated-gif-io

natari



Napari gaming
Sliding puzzle
Restore the image by reordering the superpixels using the W, A, S, D keys! 

Cell counting arcade
Commander! Cells are intruding our dish! Control your tiny space ship using 1 and 2 keys to move it left/right.
Use the 9 key to shoot a anti-cell bullet.

The image originates from BBBC022v1 (Gustafsdottir et al., PLOS ONE, 2013), available from the Broad Bioimage Benchmark Collection (Ljosa et al., Nature Methods, 2012).
Snake
Two mitochondria navigating in a cell searching for stress granules. 
The two players can control their mito using the W, A, S, D and I, J, K, L  keys, respectively.

Ping pong
Don't drop the organoid! Use your racket and hit it back to your colleague!
The two players can use W, S and I, K to control their racket, respectively.

This napari plugin was generated with Cookiecutter using with @napari's cookiecutter-napari-plugin template.
Installation
You can install natari via pip:
pip install natari

Known issues

To make the keyboard buttons work, you sometimes have to click within the image after starting the game.

Contributing
Contributions are very welcome. 
License
"natari" is free and open source software. The code is in the public domain.
See also: unlicense.org
Issues
If you encounter any problems, please file an issue along with a detailed description.

natari

Meizhu Liang

napari-generic-SIMulator






A napari plugin to simulate raw-image stacks of Structured illumination microscopy (SIM). 
The simulation is originally based on the paper GPU-accelerated real-time reconstruction in Python of three-dimensional datasets from structured illumination microscopy with hexagonal patterns by
Hai Gong, Wenjun Guo and Mark A. A. Neil (https://doi.org/10.1098/rsta.2020.0162). 
The calculation can be GPU-accelerated if the CUPY (tested with cupy-cuda11x) is installed. In addition, the TORCH package can complete the acceleration both on CPU if TORCH is installed, and on GPU if TORCH is compiled with the CUDA (tested with torch v1.13.1+cu117) enabled.
Currently applies to:
- conventional 2-beam SIM data with 3 angles and 3 phases
- 3-beam hexagonal SIM data with 7 phases, as described in the paper
- 3-beam hexagonal SIM data with 5 phases at right-angles
- conventional 3-beam 3-D data with 3 angles and 5 phases

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-generic-SIMulator via pip:
pip install napari-generic-SIMulator

To install latest development version :
pip install git+https://github.com/Meizhu-Liang/napari-generic-SIMulator.git

This plugin is compatible with napari 0.4.17 or above, older versions of napari would show errors in interpolation.
Usage
1) Open napari and create the viewer.
2) Launch two widgets: Point cloud generator and SIM data generator in Plugin.

The two widgets can be tabbed together.
   
3) Choose the type and other parameters of point cloud as a sample in Point cloud generator.
![raw](https://github.com/Meizhu-Liang/napari-generic-SIMulator/raw/main/images/pc.png)

The point cloud can be displayed in three dimensions, and be saved and loaded as .pcd files.

https://user-images.githubusercontent.com/74197598/227589232-9006842b-6706-48b7-9f2b-fe93c6698503.mp4

4) Adjust parameters in SIM data generator to simulate a raw image stack.
Apart from basic parameters such as the refractive index, the wavelengths and so on, the z scanning can be either 
   z drift: the conventional SIM (imaging a raw stack at the same z-position) or z step: the drifting case in 
   the papaer mentioned above (imaging only one raw image at a z-position).

The parameters used in the simulation can be saved with the image stack by clicking save tif with tags. Tags (of current or of one stack dragged into napari viewer) can be printed in Python by print tags. 
5) Three-dimensional point spread function (PSF), optical transfer function (OTF) and illumination patterns applied in the simulation can be showed by buttons. Note the all of these correspond the generated raw-image stack, so keep the parameters the same before showing the PSF (or OTF and illumination).
https://user-images.githubusercontent.com/74197598/227588321-ad3c8f17-1c61-4079-9e34-9b1f990714c1.mp4

https://user-images.githubusercontent.com/74197598/227586957-b76ad56e-44d5-4d9b-a1cc-2cfd08ca5400.mp4

https://user-images.githubusercontent.com/74197598/227585827-64531265-b4fb-48a9-9698-7f263f22d718.mp4

6) The raw image stacks can be then processed by napari-sim-processor (https://www.napari-hub.org/plugins/napari-sim-processor).

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-generic-SIMulator" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-generic-SIMulator

A napari plugin to simulate raw-image stacks of Structured illumination microscopy (SIM).

Sam Thiele

napari-hippo
| |  |  | 
| -------------------------- | ------------------------------- |--------------------|-------------------|
A large and slightly clumsy plugin for viewing and analysing hyperspectral data in Napari.


Installation
Follow these instructions to first install napari. Then you can install napari-hippo via pip:
pip install napari-hippo

Documentation
We are in the process of building a documentation wiki for this plugin here.
Contributing
Contributions are very welcome! Please feel free to submit pull requests or tell us about your ideas (or problems) on the discussions page.
License
Distributed under the terms of the MIT license,
napari-hippo is free and open source software.
Issues
If you encounter any problems, please file an issue along with a detailed description.
Citation
A citation for napari-hippo will be announced shortly.

napari-hippo

A fat and clumsy collection of hyperspectral tools

Maks Hess

napari-feature-classifier






An interactive classifier plugin that allows the user to assign objects in a label image to multiple classes and train a classifier to learn those classes based on a feature dataframe.
Usage

To use the napari-feature-classifier, you need to have a label image and corresponding measurements: as a csv file, loaded to layer.features or in an OME-Zarr Anndata table loaded with another plugin. Your feature measurements need to contain a label column that matches the label objects in the label image.
These interactive classification workflows are well suited to visually define cell types, find mitotic cells in images, do quality control by automatically detecting missegmented cells and other tasks where a user can easily assign objects to groups.
Prepare the label layer:

Load your label layer into napari and add the features measurements to layer.features of the corresponding label layer. You can have multiple label layers with their features open at the same time
To load features from a CSV file: Plugins -> napari-feature-classifier -> CSV Feature Loader, then load the features for the correct label image.
To load features from an OME-Zarr file: Get both the label layer into memory as a normal label layer (not a pyramidal label layer, currently untested) and the corresponding features. If your OME-Zarr file is created by Fractal, you can use this ROI loader plugin.
To load features from anywhere else, load them manually to your label_layer.features


Your feature table should have 2 columns used for indexing (but stored as normal columns in layer.features):
The label column to match the object in the label image
The roi_id column to identify the image you're currently classifying (used when a classifier is trained on multiple label images)



Initialize a classifier:

Start the classifier in napari by going to Plugins -> napari-feature-classifier -> Initialize a Classifier 
Select the features you want to use for the classifier (you need to do the feature selection before initializing. The feature selection can't be changed after initialization anymore). Hold the command key to select multiple features. Feature options are always shown for the features available in the last selected label layer, based on layer.features available features.
(Optional) Give your classes recognizable names (e.g. Mitotic & Interphase, Cell Type a, b and c etc.)


Classify objects:


Make sure you have the label layer selected on which you want to classify
Select the current class with the radio buttons or by pressing 0, 1, 2, etc.
Click on label objects in the viewer to assign them to the currently selected class
Once you have trained enough examples, click "Run Classifier" to run the classifier and have it make a prediction for all objects. Aim for at least a dozen annotations per class, as the classifier divides your annotations 80/20 in training and test sets. 
Once you get predictions, correct mistakes the classifier made and retrain it to improve its performance.
You can save the classifier under a different name or in a different location. Define the new output location and then click Save Classifier (you need to click the Save Classifier button. Just defining the new output path does not save it yet. But every run of the classifier triggers an autosave)


Apply the classifier to additional images:

You can apply a classifier trained on one image to additional label images. Use Plugins -> napari-feature-classifier -> Load Classifier 
Select the classifier (.clf file with the name you gave above) while already having the label images ready (see Prepare the label layer above).
Click Load Classifier, proceed as above.


Export classifier results

To export the training data and the results of the classifier, define an Export Name (full path to an output file or just a filename ending in .csv) where the results of the classifier shall be saved. It defaults to the layer name for the selected layer in the last directory you chose (or the current working directory if none was chosen so far)
Click Export Classifier Result (Just selecting a filename is not enough, you need to click the export button). This will export the predictions for the currently selected layer.
The results of the classifier are save in a csv file. The label is an integer of the label object within that image. The prediction column contains predictions of the classifier for all objects (except those that contained NaNs in their feature data) and the annotation column contains the annotations you made (NaN for unclassified objects, -1 for objects you deselected, 1 - 9 for the classes)


Batch mode result export
(To be updated: Create a new notebook to run batch processing, this is for the older version of the classifier)
There is a simple workflow for the classifier in the examples folder:
- Install jupyter-lab (pip install jupyterlab)
- Open the notebook in jupyter lab (Type jupyter-lab in the terminal when you are in the examples folder)
- Follow the instructions to generate an example dataframe and an example label image
- Use the classifier in napari with this simplified data
Initializing the Annotator
You can use the annotation functionality also independently from the classifier
Start the annotator widget by going to Plugins -> napari-feature-classifier -> Annotator
Select names for your classes. You can name up to 9 classes. Only classes that you give a name will be created upon initialization.
Then click Initialize.

A annotator widget opens. Use the Radio-Buttons to select what class you're annotating (or keybindings for 1-9 for classes, 0 for deselect).
The annotator will always work on the currently selected label layer. While the annotator is open, you can't edit the labels. Restart napari to allow editing of labels again.

The annotations are saved in the layer.features table of the corresponding label layer as an annotations column.

Installation
This plugin is written for the new napari npe2 plugin engine. Thus, it requires napari >= 0.4.13.
Activate your environment where you have napari installed (or install napari using pip install "napari[all]"), then install the classifier plugin:
pip install napari-feature-classifier

The layer.features dataframes have some issues in napari 0.4.17 (see here). They seem to be working again in the nighlty builds. To set up a nightly builds napari env, do the following:
conda create -n classifier-dev-napari-main -c "napari/label/nightly" -c conda-forge napari python=3.10 -y
Similar napari plugins
If you're looking for other classification approaches, apoc by Robert Haase has a pixel classifier in napari and an object classification workflow:
napari-accelerated-pixel-and-object-classification (APOC)
Alternatively, Clément Cazorla has built napari-svetlana, a deep learning based classifier
Release process

Update the version number in src/napari-feature-classifier/init.py
Update the version in setup.cfg
Add a Github release with a new version tag (matching the version set above)
Once tests pass, this should automatically be deployed to pypi
Wait for conda automation to make a PR for an updated conda release (see https://github.com/conda-forge/napari-feature-classifier-feedstock). This can take 1-2 days. Make sure that PR gets merged.

Contributing
Contributions are very welcome.
License
Distributed under the terms of the BSD-3 license,
"napari-feature-classifier" is free and open source software
Issues
If you encounter any problems, please [file an issue] along with a detailed description.
Contributors
Joel Lüthi & Max Hess

napari-feature-classifier

An interactive classifier plugin to use with label images and feature measurements

giovanni palla


napari-spatialdata: interactive exploration and annotation of spatial omics data








This repository contains a napari plugin for interactively exploring and annotating
SpatialData objects. Here you can find the [napari-spatialdata documentation]
(https://spatialdata.scverse.org/projects/napari/en/stable/notebooks/spatialdata.html). napari-spatialdata is part of the SpatialData ecosystem. To learn more about SpatialData, please see the spatialdata documentation.
Installation
You can install napari-spatialdata via pip:
pip install napari-spatialdata[all]

The all command will install the qt bindings PyQt5.
You can find more details on this in the installation instructions.
Using napari-spatialdata as default zarr reader
If you would like to use the plugin as the default zarr reader, in napari please go to File -> Preferences
-> Plugins and follow the instructions under File extension readers.
Development Version
You can install napari-spatialdata from Github with:
pip install git+https://github.com/scverse/napari-spatialdata

Or, you can also install in editable mode after cloning the repo by:
git clone https://github.com/scverse/napari-spatialdata
cd napari-spatialdata
pip install -e .

Note: when performing an editable install of napari-spatialdata, spatialdata
will be reinstalled from pip. So, if you previously also made an editable install
of spatialdata, you need to re-run pip install -e . on the spatialdata
repository. Please find more details on this in the installation instructions.
Getting started
To learn how to use the napari-spatialdata plugin, please see the documentation.
To learn how to integrate napari-spatialdata into your analysis workflows, please
see the SpatialData tutorials. In particular:

Annotating regions of interest with napari
Use landmark annotations to align multiple -omics layers

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-spatialdata" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.
Citation
Marconato, L., Palla, G., Yamauchi, K.A. et al. SpatialData: an open and universal data framework for spatial omics. Nat Methods (2024). https://doi.org/10.1038/s41592-024-02212-x
napari-spatialdata is part of the scverse® project (website, governance) and is fiscally sponsored by NumFOCUS.
If you like scverse® and want to support our mission, please consider making a tax-deductible donation to help the project pay for developer time, professional services, travel, workshops, and a variety of other needs.






napari-spatialdata

Interactive visualization of spatial omics data with napari

Abigail S McGovern

Juan Nunez-Iglesias

iterseg




napari plugin for iteratively improving a deep learning-based unet-watershed segmentation. 

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
Install iterseg using pip. Assuming you have python and pip installed (e.g., via miniconda), you can install iterseg with only one line, typed into terminal (MacOS/Linux) or annaconda prompt (Windows). We recomend installing into a new environment as some of our dependencies may not play well in the sandpit with certain versions of packages that may exist in a prexisting one. 
bash
pip install iterseg napari
Opening iterseg
Once iterseg is installed, you can access it through the napari viewer, which you can open from the command line (e.g., terminal (MacOS), anaconda prompt (Windows), git bash (Windows), etc.). To open napari simply type into the command line:
bash
napari
Loading data
Once you've opened napari, you can load image, labels, or shapes data through the load_data widget. to open the widget go to plugins/iterseg/load_data at the top left of your screen (MacOS) or viewer (Windows). 

Once the widget appears at the right of the napari window, enter the name you want to give the data you are loading (this will appear in the layers pannel on the left of the window). Choose the type of layer you want to load (Image, Labels, or Shapes: segmentations are loaded as labels layer). You can load a folder of files or a zarr file using "choose directory" (zarrs are recognised as a folder of files) or you can load a tiff file using "choose file". You can tell the program what the scale of the 3D frames will be in (in the format (z, y, x)).

If you are using a single image file (3D, 4D, 5D - ctzyx) or a directory of 3D images (zyx), for "data type" choose "individual frames". If you are using a directory of 4D images (tzyx) choose "image stacks". If you are loading a file that is 4D or 5D and want to load time points (4D: tzyx, czyx) or channels (5D: ctzyx) as individual layers, select "split channels". 
Segmenting images
You can segment data using the "segment_data" widget, which can be found at plugins/iterseg/segment_data. Once the widget appears, you can choose (1) the image layer you want to segment, (2) the folder into which to save the data, (3) the name you want to give the output file, (4) the type of segmentation to use, (5: optionally) the path to a neural network or configuration file, (6: optionally) a layer produced during training which contains metadata pointing to the trained neual network, (7) chunk size (the size of the neural network input), (8) margin (the margin of overlap between chunks). There is also an optional tickbox for debugging. If this is selected, errors will be easier to identify but you won't be able to interact with the viewer until the segmentation is done. 


Segmented images can be used to more quickly generate ground truth for training, to assess segmentation quality, or for downstream analyses. 
Segmentation algorithms
Affinity U-Net Watershed
The affinity U-net watershed is a feature based instance segmentation algorithm. A trained U-net predicts an edge affinity graph (basically boundaries in the x, y, and z axes), a map of centre points, and a mask that specifies which pixesl belong to objects. The feature map is fed to a modified watershed algorithm. The object centres are used to find seeds for the watershed and the affinity graph is used to find bounaries between objects. If you train a network using iterseg, you can select the outputted network file to segment. Otherwise, if one is not selected, a network we have trained to detect platelets will be used. This might be appropriate for small objects with high anisotropy. 
DoG Blob Segmentation
The DoG blob segmentation uses a difference of Gaussian (DoG) filter to find blob shaped objects. The DoG filter is used to find object seeds, a foreground mask, and is fed to a watershed to label objects. This algorithm cannot be trained but can be configured with a configuration file. An example configuration file can be seen in the example folder in this repository. Please see the Segmentation_config.md file for more details. 
Generating ground truth
We include two tools that are useful for generating ground truth: "save frames" and "ground truth from ROI". 
Save frames
The first tool is "save frames" can be found at plugins/iterseg/save_frames. It enables you to save frames of interest from a  series of segmented images or timeseries. 

Ground truth from ROI
The "ground truth from ROI" tool can be found at plugins/iterseg/ground_truth_from_ROI. This tool enables you to take a small portion of corrected data and place it into a new frame, which can be used for training. The new data can be tiled in the new frame to overrepresent the data in the training data set. At present, the ROI must be selected by adding a shapes layer (added using the icon circled in orange), then adding a rectangle (blue circle).

The rectangle will be used to select a region of the xy-plane. This can be seen in 3D below. 

At present, the entire z stack above and below the rectangle will be used to generate ground truth. We aim to incorporate 3D bounding boxes in the future. If multiple ROIs are selected, multiple new image frames will be made, each with a single ROI. When you generate ground truth from the shapes layer, you are able to select the desired shapes layer, image layer, and labels layer. Additionally, you can choose how many times you want to tile the ROI and how much padding to leave between. Tiling will start at the top right and progress right before moving to the next row. You will also be able to choose the save name and the folder into which to save the data. 

Training a network
iterseg includes a widget for training a u-net for the u-net affinity watershed. The training widget can be found at plugins/iterseg/train_from_viewer. Before training, you will need to load the images and ground truth you want to train from. The images and ground truth should each be a series of 3D frames that are stacked into a layer (we suggest loading from a series of frames in a directory). Once loaded, you are able to select a layer as the ground truth and a layer as the image data. You can tell the program what the scale of the output frames will be (in the format (z, y, x)). You can select what type of center prediction to use (we suggest centredness), what type of prediction to use for the mask, and what extent of affinities you want to train (if n = 1, the network will predict only the direct boundaries between objects in each axis, if greater than 1 the network will still predict the direct boundaires but will also predict where there is a new object n steps away - can be used as collateral learning to enhance training). Affinities extent is developmental. Please submit an issue for any problems. 

For the U-net training, we allow you to choose the learning rate for the ADAM optimiser used to train the network. You can also choose between binary cross entropy loss (BCELoss) and Dice loss (DICELoss). We have found in our data that BCE loss works better. You can also choose how many chunks of data are produced from each frame (n each) and how many epochs you want to train for the training will be done in n_each * n_frames batches with a minibatch size of 1. 
In the future we hope to expand this training widget to enable training other types of networks. Please get involved if you feel you can help with this. 
Assessing segmentations
iterseg includes widgets for assessing and comparing segmentations. If you want to assess segmentation quality, you will need to load a ground truth and a segmentation to assess. Once loaded, you can select the ground truth and segmentation (model segmentation) using the widget found in plugins/iterseg/assess_segmentation. You can select which metrics you want to assess. The metrics we enable are:
- Variation of information (VI): VI is a two part measure. It includes a measure of undersegmentation and oversegmentation. Undersegmentation is a measure of the amount of new information you get from looking at the ground truth if you have already seen the segmentation. It can be interpreted as the proportion of objects that are incorrectly merged. Oversegmentation is a measure of the amount of new information you get from looking at the segmentation if you have already seen the ground truth. It can be interpreted as the proportion of objects that are incorrectly split. For more info please see the scikit-image documentation. 
- Object count difference (OD): Object count difference is simply the difference in number of objects between a ground truth and the assessed segmentation (card(ground truth) - card(segmentation)). 
- Average precision (AP): Average precision  Average precision is a combined measure of how accurate the model is at finding true positive (real) objects (we call this precision) and how many of ground truth real objects it found (this is called recall). The assessment of whether an object is TP, FP, and FN depends on the threashold of overlap between objects. Here we use the intersection of union (IoU), which is the proportion of overlap between the bounding boxes of ground truth and model segemented objects. AP is assessed using different IoU thresholds (from 0.35-0.95). The resultant data will be plotted as IoU by AP. 

Precision = TP / (TP + FP)Recall = TP / (TP + FN). 
Abbreviations: FN, false negative; TP, true positive; FP, false 

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"iterseg" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

iterseg

napari plugin for iteratively improving unet-watershed segmentation

Patrick Cleeve

napari-openfibsem






OpenFIBSEM Applications

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-openfibsem via pip:
pip install napari-openfibsem

To install latest development version :
pip install git+https://github.com/DeMarcoLab/napari-openfibsem.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-openfibsem" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-openfibsem

Da Kuang

napari-mask-density








A napari plugin for analyzing mask density in regions of interest

This napari plugin was generated with copier using the napari-plugin-template.

Installation
You can install napari-mask-density via pip:
pip install napari-mask-density

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-mask-density" is free and open source software
Issues
If you encounter any problems, please [file an issue] along with a detailed description.

napari-mask-density

A napari plugin for analyzing mask density in regions of interest

Piers Turner

napari-molSEEQ




A Napari plugin for extracting single molecule sequences from single/multi-channel SMLM microscopy data. 
Compatible with both ALEX and FRET data. All functions are parallelised/GPU accelerated where possible to increase performance.
Multiple datasets can be loaded and processed in parallel.
napari-molseeq uses Picasso (picassosr) as a backend and includes features for aligning image channels/datasets, undrifting images, detecting/fitting localisations and extracting traces, and supports both ALEX and FRET data. Traces can be exported in different formats for downstream analysis.
napari-molseeq traces can be analysed with TraceAnalyser: https://github.com/piedrro/TraceAnalyser
This is still undergoing development, so some features may not work as expected.
This was built by Dr Piers Turner from the Kapanidis Lab, University of Oxford.

Installation
You can install napari-molseeq via pip:
pip install napari-molseeq

You can install napari-molseeq via [GitHub]:
conda create –-name napari-molseeq python==3.9
conda activate napari-molseeq
conda install -c anaconda git
conda update --all

pip install git+https://github.com/piedrro/napari-molseeq.git

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-molseeq" is free and open source software
Issues
If you encounter any problems, please file an issue along with a detailed description.

napari-molseeq

A Napari plugin for extracting single molecule sequences from single/multi-channel SMLM microscopy data.

Supported file formats:
 - Fluidigm MCD
 - Fluidigm TXT
Supported image types:
  - Panoramas (single-channel, color)
  - Acquisitions (multi-channel, grayscale)
All images are loaded co-registered within the machine's coordinate system.

napari-imc

Imaging Mass Cytometry (IMC) file type support for napari

NeuroFly: A framework for single neuron reconstruction at whole-brain scale
This package provides tools for semi-automatic neuron reconstruction. Features based on deep learning, like image segmentation and deconvolution are implemented in tinygrad, which can run on almost any GPU (NVIDIA, AMD, Apple, Qualcomm, Intel).

Update
Our transformer-based autonomous driving model is available now!
Press shortcut 'd' to automatically extend segments.

Installation
Install the latest version
pip install --upgrade git+https://github.com/beanli161514/neurofly.git
or
pip install neurofly
You can also install from napari hub, using their plugin manager with GUI.
Dataset
We provide several expert-proofread reconstruction results for testing, model training, and evaluation. Zenodo Link
Content of samples
| name           | size  | species | label type  | imaging |
|----------------|-------|---------|-------------|---------|
| rm009_labeled  | 629MB | macaque | skeleton    | VISoR   |
| mouse_labeled  | 260MB | mouse   | skeleton    | VISoR   |
| z002_labeled   | 204MB | mouse   | skeleton    | VISoR   |
| fmost_labeled  | 370MB | mouse   | skeleton    | fMOST   |
| RM009_noisy_1  | 65MB  | macaque | morphology  | VISoR   |
| RM009_noisy_2  | 65MB  | macaque | morphology  | VISoR   |
| fmost_test     | 65MB  | mouse   | morphology  | fMOST   |
| z002_dendrites | 768MB | mouse   | morphology  | VISoR   |
| RM009_arbor_1  | 288MB | macaque | morphology  | VISoR   |
| RM009_axons_1  | 600MB | macaque | morphology  | VISoR   |
| RM009_axons_2  | 600MB | macaque | morphology  | VISoR   |
| z002           | 8.92G | mouse   | morphology* | VISoR   |
$*$ annotation in progress
Label format
Morphology labels are graphs saved in SQLite database with 3 tables:
|    segments    |  nodes  |  edges  |
|:--------------:|:-------:|:-------:|
|       sid      |   nid   |   src   |
|     points     |  coord  |   des   |
| sampled_points | creator |   date  |
|                |  status | creator |
|                |   type  |         |
|                |   date  |         |
|                | checked |         |
Segments are results of the segmentation stage, they are used to generate initial nodes and edges.
Basic usage example

NeuroFly packaged 4 napari plugins for image browsing, image segmentation, and data annotation.

Segmentation
NeuroFly supports whole brain image saved in hierarchical data structures(ims, h5, and zarr) in Imaris File Format and small image volumes saved in single-channel tiff format. Here we use a mouse brain in our dataset named z002.zarr.zip as example.
This brain is sparsely labeled, which means only a tiny puny part of neurons are lighted and imaged. To extract these foreground singals, you can use the provided command line interface 'seg'. We provide a default weight trained on images captured by VISoR and fMOST.
seg -i z002.zarr.zip -vis -d z002.db
or use the graphical interface

This process may take about 10 hours depending on your you hardware configuration. When finished, you should see the extracted segments and a database file named z002.db in your working dictionary.
An image block with severe contamination and the segmentation result

Manual connection and proofreading
Launch annotation tool from napari menu, Plugin -> neurofly -> Segs Annotator
Load data
Load image file (z002.zarr.zip) and database file (z002.db), then click refresh panorama button to show the panorama view.

Select one node as start point of annotation
In panorama mode, you can easily identify sparse, bright signals that are promising for reconstruction. The silde bars 'short segs filter', 'length thres', and 'point size' can be adjusted to hide noise and short segments. 
If you can clearly identify foreground segments, click on one of the conspicuous segments to select it as start point of annotation. Once selected, the id of picked node will be displayed at node selection. Then click 'switch mode' to switch to labeling mode, and the tasks will be generated automatically.
Task generation
Given a selected node, task generator analyses its connected component and extract all unchecked terminal nodes. The tasks are designed very simple: Connect the center node with the surrounding nodes if there should be an edge. The criterion is whether the edge aligns well with the imaged neuron fibers.

Node operations

In each task, a center node and nearby segments are rendered, you can add/remove nodes and edges to get a reasonable local structure.
Left click on nodes to add/remove an edge between it and the center node

Right click to remove a node

Press 'g' or use left panel to switch to 'image' layer, then right click to add points

Use dropdown selection in right panel to add type label for center node.

Deconvolution
Press 'i' or click on 'deconvolution' to deconvolve the image

Proofreading
If you find something wrong when labeling, for example, two somas are connected together. You can use proofreading mode to check the neuron branch by branch.

Performance
NeuroFly supports rendering of more than ten million points. (tested on M3 Macbook Air and RTX 3090 workstation)

Export as swc file
Switch to panorama mode, adjust 'length_thres' to filter out short segments and keep only complete neurons. Then press 'export swc files', each neuron will be saved as one .swc file in your working dictionary.


neurofly

A framework to annotate single neurons at whole-brain scale

christinab12

Napari Organoid Counter - Version 0.2 is out!








A napari plugin to automatically count lung organoids from microscopy imaging data. Note: this plugin only supports single channel grayscale images.
Hold it for the demo!


This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.
Installation
This plugin has been tested with python 3.9 and 3.10 - you may consider using conda to create your dedicated environment before running the napari-organoid-counter.


You can install napari-organoid-counter via pip:
pip install napari-organoid-counter


To install latest development version :
```pip install git+https://github.com/HelmholtzAI-Consultants-Munich/napari-organoid-counter.git```



Additionally, you will then need to install one additional dependency:
mim install "mmcv<2.2.0,>=2.0.0rc4"


For installing on a Windows machine directly from within napari, follow the instuctions here. Step 2 additionally needs to be performed here too (mim install "mmcv<2.2.0,>=2.0.0rc4").
What's new in v2?
Checkout our What's New in v2 here.
How to use?
After installing, you can start napari (either by typing napari in your terminal or by launching the application) and select the plugin from the drop down menu.
For more information on this plugin, its' intended audience, as well as Quickstart guide go to our Quickstart guide.
Contributing
Contributions are very welcome. Tests can be run with [pytest], please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the MIT license,
"napari-organoid-counter" is free and open source software
Dependencies
napari-organoid-counter uses the napari-aicsimageio[1] [2] plugin for reading and processing CZI images.
[1] Eva Maxfield Brown, Dan Toloudis, Jamie Sherman, Madison Swain-Bowden, Talley Lambert, AICSImageIO Contributors (2021). AICSImageIO: Image Reading, Metadata Conversion, and Image Writing for Microscopy Images in Pure Python [Computer software]. GitHub. https://github.com/AllenCellModeling/aicsimageio
[2] Eva Maxfield Brown, Talley Lambert, Peter Sobolewski, Napari-AICSImageIO Contributors (2021). Napari-AICSImageIO: Image Reading in Napari using AICSImageIO [Computer software]. GitHub. https://github.com/AllenCellModeling/napari-aicsimageio
The latest version also uses models developed with the mmdetection package [3], see here
[3] Chen, Kai, et al. "MMDetection: Open mmlab detection toolbox and benchmark." arXiv preprint arXiv:1906.07155 (2019).
Issues
If you encounter any problems, please file an issue along with a detailed description.
Citing
If you use this plugin for your work, please cite it using the following:

Christina Bukas, Harshavardhan Subramanian, & Marie Piraud. (2023). HelmholtzAI-Consultants-Munich/napari-organoid-counter: v0.2.0 (v0.2.0). Zenodo. https://doi.org/10.5281/zenodo.7859571
bibtex:
@software{christina_bukas_2022_6457904,
  author       = {Christina Bukas, Harshavardhan Subramanian, & Marie Piraud},
  title        = {{HelmholtzAI-Consultants-Munich/napari-organoid- 
                   counter: second release of the napari plugin for lung
                   organoid counting}},
  month        = apr,
  year         = 2023,
  publisher    = {Zenodo},
  version      = {v0.2.0},
  doi          = {10.5281/zenodo.7859571},
  url          = {https://doi.org/10.5281/zenodo.7859571}
}


napari-organoid-counter

A plugin to automatically count lung organoids using Deep Learning.

napari-assistant








The napari-assistant is a napari meta-plugin for building image processing workflows. 
Usage
After installing one or more napari plugins that use the napari-assistant as user interface, you can start it from the 
menu Tools > Utilities > Assistant (na) or run naparia from the command line. 
By clicking on the buttons in the assistant, you can setup a workflow for processing the images.

While setting up your workflow, you can at any point select a layer from the layer list (1) and change the parameters of
the corresponding operation (2). The layer will update when you change parameters and also all subsequent operations. 
You can also vary which operation is applied to the image (3). Also make sure the right input image layer is selected (4).

Saving and loading workflows
You can also save and load workflows to disk. 

After loading a workflow, make sure that the right input images are selected.
Code generation
The napari-assistant allows exporting the given workflow as Python script and Jupyter Notebook. 

Furthermore, if you have the napari-script-editor installed,
you can also send the current workflow as code to the script editor from the same menu.

Plugin generation
There is also a Napari plugin generator available. Check out its documentation to learn how napari-assistant compatible plugins can be generated directly from within the assistant.
Installation
It is recommended to install the napari-assistant via one of the plugins that use it as graphical user interface.
You find a complete list of plugins that use the assistant on the napari-hub.
Multiple of these plugins come bundled when installing devbio-napari.
Note: This plugin is not compatible with napari 0.6.0 and later.
For developers
If you want to make your napari-plugin accessible from the napari-assistant, consider programming functions with a simple 
interface that consume images, labels, integers, floats and strings. Annotate input and return types, e.g. like this:
```python
def example_function_widget(image: "napari.types.ImageData") -> "napari.types.LabelsData":
    from skimage.filters import threshold_otsu
    binary_image = image > threshold_otsu(image)
from skimage.measure import label
return label(binary_image)

```
Furthermore, please add your function to the napari.yaml which uses npe2:
name: napari-npe2-test
display_name: napari-npe2-test
contributions:
  commands: 
    - id: napari-npe2-test.make_magic_widget
      python_name: napari_npe2_test._widget:example_magic_widget
      title: Make example magic widget
  widgets:
    - command: napari-npe2-test.make_magic_widget
      display_name: Segmentation / labeling > Otsu Labeling (nnpe2t)
To put it in the right button within the napari-assistant, please use one of the following prefixes for the display_name:
* Filtering / noise removal >
* Filtering / background removal >
* Filtering >
* Image math >
* Transform >
* Projection >
* Segmentation / binarization >
* Segmentation / labeling >
* Segmentation post-processing >
* Measurement >
* Label neighbor filters >
* Label filters >
* Visualization >
You find a fully functional example here.
Last but not least, to make your napari-plugin is listed in the napari-hub when searching for "napari-assistant", make sure
you mention it in your readme.
Feedback welcome!
The napari-assistant is developed in the open because we believe in the open source community. Feel free to drop feedback as github issue or via image.sc
Contributing
Contributions are very welcome. Please ensure
the test coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the BSD-3 license,
"napari-assistant" is free and open source software
Acknowledgements
This project was supported by the Deutsche Forschungsgemeinschaft under Germany’s Excellence Strategy – EXC2068 - Cluster of Excellence "Physics of Life" of TU Dresden. 
This project has been made possible in part by grant number 2021-240341 (Napari plugin accelerator grant) from the Chan Zuckerberg Initiative DAF, an advised fund of the Silicon Valley Community Foundation.

napari-assistant

A pocket calculator like interface to image processing in napari

Karol Gotkowski

napari-blosc2




An image reader & writer for blosc2 images.

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation
You can install napari-blosc2 via pip:
pip install napari-blosc2

Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
Distributed under the terms of the Apache Software License 2.0 license,
"napari-blosc2" is free and open source software
Issues
If you encounter any problems, please [file an issue] along with a detailed description.

napari-blosc2

An image reader & writer for blosc2 images.

Qiqi Lu

napari-kld




napari-kld is a napari plugin that implements kernel learning deconvolution algrotihm.
Kernel Learning Deconvolution (KLD)
KLD is a rapid deconvolution algorithm for fluorescence microscopic image, which learns the forward and backward kernels in Richardson-Lucy Deconvolution (RLD) from paired low-resolution (LR) and high-resolution (HR) images.
It only requires one sample to training the model, and two iterations to achieve a superior deconvolution performance compared to RLD and its variants using unmatched backward projection.
*This napari plugin was generated with copier using the napari-plugin-template.
Installation
You must install napari firstly and then install napari-kld.
Install napari
You can download the napari bundled app for a simple installation via https://napari.org/stable/tutorials/fundamentals/quick_start.html#installation.
Or, you can install napari with Python using pip:
conda create -y -n napari-env -c conda-forge python=3.10
conda activate napari-env
python -m pip install 'napari[all]'
Refer to https://napari.org/stable/tutorials/fundamentals/quick_start.html#installation.
Install napari-kld
You can install napari-kld plugin with napari:
Plugins > Install/Uninstall Plugins… > [search napari-kld] > install.
Or you can install napari-kld via pip:
pip install napari-kld

Instruction
This plugin includes two part:


RL Deconvolution : Conventional RLD algorithm using different type of backward kernels (including matched backward kernel [Traditional] and unmatched backward kernels [Guassian, Butterworth, Wiener-Butterworth (WB)]). The forward kernel, i.e., point spread function (PSF), is required.


KL Deconvolution : KLD using learned forward/backward kernels.


You can download the "test" folder  at https://github.com/qiqi-lu/kernel-learning-deconvolution for testing, which save some 2D/3D images used for training and testing.
RL Deconvolution
The conventional RLD using different type of backward kernels.


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > RL Deconvolution


Load input low-resolution (LR) image: File > Open File(s) > [choose the image to be deconvolved] > [the image will appear in the layer list of napari], such as the simulated image "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/train/raw/0.tif".


Choose the name of loaded image in Input RAW data, such as "0".


Press Choose to choose a PSF correspongding to the loaded image, such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/train/psf.tif".


Choose the type of backward kernel in Method combo box:

Traditional : the backward kernel is just the flip of forward kernel (i.e., PSF).
Guassian : Guassian-shaped backward kernel, thw FWHM of which is same as the forward kernel.
Butterworth : Butterworth-shaped backward kernel, which is constructed using Butterworth filter.
WB : WB-shaped backward kernel, which is constructed by combining Wiener and Butterworth filter.



Set the number of RL iterations Iterations and parameters of backward kernel*.


Press run button to do deconvolution.


Wait the progress bar to reach 100%.


The output deconved image will appear in the layer list named as {name of input image}_deconv_{Method}_iter_{Iterations}, such as "0_deconv_traditional_iter_30".


*The adjustment of parameters of backward kernels should refer to the paper : Guo, M. et al. Rapid image deconvolution and multiview fusion for optical microscopy. Nat Biotechnol 38, 1337–1346 (2020).
KL Deconvolution
Training data preparation
The data used for training must be prepared in a folder consisting of:

A folder named "gt" (optional) , such as "test/data/real/2D/train/gt", which saves all the GT images (only support .tif file).
A folder named "raw", such as "test/data/real/2D/train/raw", which saves all the LR images (only support .tif file). The file names must be the same as those in "gt" folder.
A file named "train.txt", such as "test/data/real/2D/train/train.txt", which saves the name of each image in "gt"/"raw" filder in each line.

When yuo have paired LR image and HR image
When we have paired LR image and HR image, we can treat LR image as raw input and HR image as ground truth (GT). We can first learn the forward kernel and then learn the backward kernel in a supervised strategy.
Training of Forward Projection
Train the forward projection to learn forward kernel.


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > KL Deconvolution


Choose Training tab.


Choose Data Directory, such as "test/data/real/2D/train". Then the dimention of training data will show in the Dimension box.


Choose Output Directory, such as "test/data/real/2D".


PSF Directory is not required as the PSF is unknown.


If the raw input and GT image have different intensity, please check the Preprocess check box, which will rescale the input and GT images to have the same intensity. Here, do not check.


In the Forward Projection box, set the parameters of training:

Epoch : number of epochs of training.
Batch Size : batch size of training data used during training.
Kernel Size (z, xy) : the size of forward kernel to learned.
Optimizer : the optimization algorithm. Default: Adam.
Learning Rate : learning rate of training.
Decay Step ： the decay step of learning rate. Note: 0 for no decay.
Decay Rate : the decay rate of learning rate.



Press run button. You can press the stop button to end the training.


Wait the progress bar to reach 100% and training finished.


After the training of forward projection, the results will be save in the /checkpoints folder in Output Directory, the model was named as forward_bs_{batch size}_lr_{learning rate}_ks_{kernel size (z)}_{kernel size (xy)}, such as "test/data/real/2D/checkpoints/forward_bs_1_lr_0.001_ks_1_31", which consists of:
- a log folder saved the Tensorboard log, which can be opened with Tensorboard.
- many model checkpoints, named as epoch_{epoch}.pt.
- a parameters.json file saving the parameters used to training the model.
Training of Backward Projection
After training of forward projeciton, we can freeze the forward projeciton and then train the backward projeciton.


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > KL Deconvolution


Choose Training tab.


Choose Data Directory, such as "test/data/2D/real/train". Then the dimention of training data will show in the Dimension box.


Choose Output Directory, such as "test/data/2D/real".


PSF Directory is no required as the PSF is unknown.


If the raw input and GT image have different intensity, please check the Preprocess check box, which will rescale the input and GT images to have the same intensity. Here, do not check.


In the Backward Projeciton box, set parameters for the trianing of backward projeciton.

Training strategy : supervised training or self-supervised training. Here, set as supervised, as we have the GT images. When choosing the self-supervised mode, a PSF is required.
Iterations (RL) : The number of iterations of RL iterative procedure. Default: 2.
Epoch : The number fo epochs used to traing the model.
Batch Size : The batch size used to training the model.
Kernel Size (z, xy): The size of backward kernel, x and y have the same size.
FP directory : the directory of the pre-trained forward projeciton model, such as "test/data/real/2D/checkpoints/forward_bs_1_lr_0.001_ks_1_31/epoch_500_final.pt" (commonly the model labeled with "_final" is used).
Optimizer : Optimization algorithm. Default: Adam.
Learning Rate : The learning rate used to trianing the model.
Decay Step : the decay step of learning rate.
Decay Rate : the decay rate of learning rate.



Press run button. You can press the stop button to end the training.


Wait the progress bar to reach 100% and then the training finishes.


When the training finishes, the results will be save in the /checkpoints folder in Output Directory, the model was named as backward_bs_{batch size}_lr_{learning rate}_iter_{num of RL iterations}_ks_{kernel size (z)}_{kernel size (xy)}, such as "test/data/real/2D/checkpoints/backward_bs_1_lr_1e-05_iter_2_ks_1_31", which consists of:

a log folder saved the Tensorboard log, which can be opened with Tensorboard.
many model checkpoints, named as epoch_{epoch}.pt.
a parameters.json file saving the parameters used to training the model.

Now we get the learned forward projection and backward projection.
Next, we can use them to do Prediction.
When you only have a PSF
When you only have a PSF to do deconvolution, you can train the model using simulated data following the below steps:

Generate simulaiton data.
Train the model under supervised mode.
Apply the trained model on real data.

Simulation data generation


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > KL Deconvolution


Choose Simulation tab.


Choose the Output Directory of the generated simulation data, such as "test\data\simulation".


Choose the PSF Directory (only support 2D/3D PSF file save as .tif, axes = (y, x) or (z, y, x)), such as "test\data\simulation\psf.tif".


Adjust the parameters as needed.

Image Shape : the shape of simulated image, when z=1, 2D image will be generated.
PSF Crop : when the input PSF is too large, you can crop the PSF to acuqire a smaller PSF. All the PSF will be converted to have an odd shape and normalized.
Num of Simulations : number of generated images.
Gaussian (std) : the standard deviation (std) of Gaussian noise added in the generated LR images. The mean of Gaussian noise = 0. Default: 0 (i.e., without gaussian noise).
Poisson : whether to add Poisson noise, if True, make the Enable checked.
Ratio : a ratio factor multiplied on GT image to control the level of Poisson noise, thus the simulated raw input LR image RAW can be expressed as:

$$ RAW = Possion((GT \cdot Ratio)\times PSF) + Gaussian $$

Scale Factor : downsampling scale factor. Default: 1.



Press run button.


Wait the progress bar to reach 100%.


The generated simulation data will be save in Output directory, named as "data_{shape_z}_{shape_y}_{shape_x}_gauss_{std of Gaussian noise}_poiss_{whether to add Poisson noise}_ratio_{Ratio}", such as: "test\data\simulation\data_128_128_128_gauss_0.0_poiss_0_ratio_1.0\train"

"data\train\gt" saves the GT images which consist of structures with various shapes*.
"data\train\raw" saves the RAW images with blur and noise.
"data\train\parameters.json" is a dictionary of parameters used to generate the simulation data.
"data\train\psf.tif" is the PSF used in the simulation data generation (as the original PSF may be cropped).
"data\train\train.txt save all the image used to train the model.

After you generate simulation data, you can use them to train the model.
*the code was refered to the paper: Li, Y. et al. Incorporating the image formation process into deep learning improves network performance. Nat Methods 19, 1427–1437 (2022).
You may need to adjust the noise level in the image accordding to the real acuqired data.
Training with known PSF and simulated data
The simulated data should be those generated using the known PSF.


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > KL Deconvolution


Choose Training tab.


Choose Data Directory, such as test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/train", which saves the data used to train the model in should include:

A gt folder saves the GT images
A raw folder save the raw input LR images with the same file name as GT images
A train.txt file saves all the file names used to train the model (does not need to list all the files in gt/raw folder but at least one).



Choose a Output Directory to save the model checkpoints, such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0".


Choose PSF Directory of the PSF corresponding to the data, such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/train/psf.tif". Then the Forward Projection group box will be invisible as we do not need to learn the forward kernel when we know the PSF. Just use the PSF as the forward kernel.


If the raw input and GT image have different intensity, please check the Preprocess check box, which will rescale the input and GT images to have the same intensity. Here, do not check.


Then set parameters to learn the backward kernel.

Training Strategy : supervised training or self-supervised training. Here, set as supervised, as we have the GT images.
Iterations (RL) : the number of iterations of RL iteration procedure. Default: 2.
Epoch : the number fo epochs used to traing the model.
Batch Size : the batch size used to training the model.
Kernel Size (z, xy): the size of backward kernel, x and y directions have the same size.
FP Directory : the directory of the forward projeciton model. Here, it is disabled as the PSF is known.
Optimizer : Optimization algorithm. Default: Adam.
Learning Rate : the learning rate used to trianing the model.
Decay Step : the decay step of learning rate.
Decay Rate : the decay rate of learning rate.



Press run button. You can press the stop button to end the training.


Wait the progress bar to reach 100% and the training finishes.


When the training finishes, a checkpoints folder will be created in Output Directory such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/checkpoints".
The models is save in /checkpoints folder, which is named as "backward_bs_{batch size}_lr_{learning rate}_iter_{num of RL iterations}_ks_{kernel size (z)}_{kernel size (xy)}", such as "/checkpoints/backward_bs_1_lr_1e-06_iter_2_ks_1_31", consists of:

A log folder saved the Tensorboard log, which can be open with Tensorboard.
Many model checkpoints, named as epoch_{epoch}.pt.
A parameters.json file saving the parameters used to training the model.

When you only have LR image and corresponding PSF
When we only have LR image and its PSF, we can traing the backward projection through supervised training using simulation data as introduced above. The plugin also provide an alternative self-supervised training stratergy to learn the backward kernel.


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > KL Deconvolution


Choose Training tab.


Choose Data Directory, such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/train".


choose Output Directory, such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0".


Choose PSF Directory, such as "test/data/simulation/data_128_128_128_gauss_0.0_poiss_0_ratio_1.0/train/psf.tif" then the Forward Projection box will be invisiable.


As there is no GT image, preprocessing is not needed. Do not check Preprocess check box.


In the Backward Projeciton box, set parameters for the trianing of backward projeciton.

Training strategy : supervised training or self-supervised training. Set as self-supervised, as we do not have the GT images.
Iterations (RL) : the number of iterations of RL iteration procedure. Default: 2.
Epoch : the number fo epochs used to traing the model.
Batch Size : the batch size used to training the model.
Kernel Size (z, xy): the size of backward kernel, x and y directions have the same size.
FP Directory : the directory of the forward projeciton model. Here, it is disabled as the PSF is known.
Optimizer : Optimization algorithm. Default: Adam.
Learning Rate : the learning rate used to trianing the model.
Decay Step : the decay step of learning rate.
Decay Rate : the decay rate of learning rate.



Press run button. You can press the stop button to end the training.


Wait the progress bar to reach 100% and training finishes.


When the training finishes, the results will be save in the /checkpoints folder in Output Directory, the model was named as backward_bs_{batch size}_lr_{learning rate}_iter_{num of RL iterations}_ks_{kernel size (z)}_{kernel size (xy)}_ss, such as "/checkpoints/backward_bs_1_lr_1e-05_iter_2_ks_31_31_ss", which consists of:

a log folder saved the Tensorboard log, which can be opened with Tensorboard.
many model checkpoints, named as epoch_{epoch}.pt.
a parameters.json file saving the parameters used to training the model.

The performance of self-supervised learning may be inferior to supervised learning according to our experiments.
Prediction
Use the learned forward/backward kernel to do deconvolution.


Open napari and load napari-kld plugin: Plugins > Kernel Learning Deconvolution > KL Deconvolution


Choose Prediction tab.


Load raw input low-resolution image through napari: File > Open File(s) > [choose the image to be deconvolved] > [the image will appear in the layer list of napari], such as "test/data/real/2D/test/raw/2.tif".


Choose the loaded image in Input RAW data box, e.g., 2.


If the PSF is known, choose the PSF directory.


If the PSF is unknown, choose the Forward Projection directory, such as "test/data/real/2D/checkpoints/forward_bs_1_lr_0.001_ks_1_31/epoch_500_final.pt" (commonly the model labeled with "_final" is used). If both the directories of PSF and Forward Projeciton is choosen, KLD will directly use the PSF selected.


Choose the Backward Projeciton directory, such as "test/data/real/2D/checkpoints/backward_bs_1_lr_1e-05_iter_2_ks_1_31/epoch_1000_final.pt"  (commonly the model labeled with "_final" is used).


Set the number of RL iterations at Iterations (RL). Default: 2.


Press run to do deconvolution.


Wait the progress bar to reach 100%.


The deconvolved image will be directly shown in the layer list of napari, named as "{input data name}_deconvo_iter_{number of RL iterations}", e.g., "16_deconv_iter_2". You can save it as needed.
Others
The log tab print the message during running.
Press clean button will clean all the text in the log box.
Notice


Currently, the plugin is runned on CPU. We have tried to run the training on GPU, but the training time did not decrease (maybe it is because the FFT-based covnlution was not optimized on GPU). We are trying to make improvements.


The training time may be very long if we set the kernel size or the number of epoches too large, especially for 3D images. Besides, it also depends on the  computation capability of your device.


Contributing
Contributions are very welcome. Tests can be run with tox, please ensure
the coverage at least stays the same before you submit a pull request.
License
MIT LICENSE
Issues
If you encounter any problems, please [file an issue] along with a detailed description.

napari-kld

Kernel learning deconvolution (KLD) is a rapid deconvolution algorithm for fluorescence microscopic image, which learns the forward and backward kernels in Richardson-Lucy Deconvolution (KLD) from paired low-/high-resolution images.

Plugins

Browse plugins: 1

NaPaM

First released:

Last updated:

Installs:

Plugin type: