Amend segmentation and track

Amend segmentation and track within napari manually.

:eyes: watch a demo video

Input data structure

Napari-amdtrk reads an input directory which includes:

• An intensity image (tif) in txyc (or txy) format

• An object mask (tif) in txy format

• An object table (csv) with following essential columns:

  • frame: time frame
  • trackId: ID of the track, starting from 1
  • Center_of_the_object_0: x coordinate
  • Center_of_the_object_1: y coordinate
  • continuous_label: the corresponding label (intensity value) of the object in the object mask (You may use skimage.measure.label to get it from a binary mask).

• A config file named config.yaml (other names are not allowed)

  Within the config file, there should be:

  • intensity_suffix: suffix of the intensity image (e.g., for foo_GFP.tif, use GFP in the config). For multiple intensity images, separate them with commas (e.g., GFP, mCherry)
  • mask_suffix: suffix of the mask image
  • track_suffix: suffix of the tracked object table
  • frame_base: index of the first frame (either 0 or 1)
  • stateCol: optional column name for the cell state (e.g., cell cycle phase) in the object table. Leave blank if the object table does not contain it

Napari-amdtrk will modify mask and track files in place. Other files are not affected.

Quick start

1. Open napari GUI.
2. File > Open folder > choose Amend segmentation and track
3. Plugins > napari-amdtrk: Amend track widget > Run
4. In layer list, select the segm layer to start editing.

Please check out the demo video here and the sample data (see below).

Sample data

To load sample data, File > Open Sample > napari-amdtrk > basic tracks or complete cell cycle tracks.

• basic tracks: simple cell tracks as essential input data.
• complete cell cycle tracks: cell tracks with additional cell cycle features.

The above operations will download data to ~/.amd_trk/_sample_data/ (~230MB). After downloading is finished, sample data will be loaded.

Notes

• Please cite this repository if using the plugin in your work (try About > Cite this repository upper right of this homepage).

• Sample data (cell track videos) have been published with pcnaDeep: a fast and robust single-cell tracking method using deep-learning mediated cell cycle profiling. We acknowledge Dr Kuan Yoow Chan and members of his lab for generating the data.

Keyboard shortcuts

• ↑ and ↓: toggle different operations

• enter: run the operation

• Available to a selected object:

  • control + 9: shrink the object mask
  • control + 0: expand the object mask

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

Installation

Please install napari GUI first:

python -m pip install "napari[all]"

You can install napari-amdtrk via pip:

pip install napari-amdtrk

License

Distributed under the terms of the MIT license, "napari-amdtrk" is free and open source software

Issues

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