Isolate Cell from Segmentation

A napari plugin to isolate single cell morphologies (e.g., neurons) from label volumes based on a user click, automatically read image scale, and export the isolated structure as TIFF and correctly scaled SWC files.

Overview

This plugin helps streamline the process of extracting individual cell structures from dense segmentations, such as those produced by deep learning models like nnUNet.

Key Features:

• Click-Based Isolation: Simply click on the soma (or any part) of the cell you want to isolate in a Napari Labels layer.
• Automatic Scale Detection: Reads ZYX scale information directly from TIFF metadata (standard tags or ImageJ metadata) and applies it to the loaded Napari layer.
• Anisotropy Awareness: Automatically populates the widget's Anisotropy fields based on the detected image scale.
• Outputs:
  • Adds the isolated cell as a new Labels layer in Napari, preserving the original scale.
  • Saves the isolated label volume as a TIFF file.
  • Saves the skeletonized structure as an SWC file with coordinates reflecting the original image's physical scale (micrometers).
• Configurable Parameters: Adjust morphological closing radius (defaults to 0 for dense segmentations) and skeleton dust threshold.

Workflow

Installation

For Users (Recommended)

pip install napari-isolate-cell

Or using uv (faster):

uv pip install napari-isolate-cell

For Developers

git clone https://github.com/serg-bg/napari-isolate-cell.git
cd napari-isolate-cell
pip install -e .[testing]

Usage

1. Launch napari and open your 3D segmentation (.tif file)
2. Open plugin: Plugins → napari-isolate-cell → Isolate Cell Arbor
3. Select your labels layer from the dropdown
4. Click "Activate Click Isolation"
5. Click any cell in the viewer to isolate it

Outputs:

• New labels layer with isolated cell
• isolated_outputs/ folder containing:
  • .tif - Isolated cell volume
  • .swc - Skeleton with physical coordinates (µm)

Parameters:

• Morphological Closing: Default 0 (increase to bridge small gaps)
• Dust Threshold: Default 100 (minimum skeleton branch size in voxels)
• Anisotropy: Auto-detected from TIFF metadata

Requirements

• Python >= 3.10
• napari
• NumPy
• scikit-image
• SciPy
• tifffile
• magicgui
• qtpy

(See pyproject.toml for specific version constraints)

Contributing

Contributions are very welcome. Please file an issue to discuss potential changes or features first. Tests can be run with pytest (pip install -e .[testing] then 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-isolate-cell" is free and open source software

Issues

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