napari-easytrack 0.1.4

Cell tracking with btrack presets and parameter optimization

napari-easytrack

napari-easytrack is a napari plugin for automated parameter tuning in cell tracking. It optimises btrack obtaining a set of optimal tracking parameters for a dataset. Using that optimal set of parameters, napari-easytrack can then track the cells in the dataset, improving tracking accuracy and reducing manual correction time.

napari-easytrack provides two widgets in napari:

1. An optimization widget that optimizes tracking parameters based on a small subset of manually annotated ground-truth data.
2. A tracking widget that uses the optimized parameters to track the entire dataset. Here, we provide different tracking presets so users can choose the one that best fits their data without optimization. If no preset fits the data, users should try to optimize the parameters first with the optimization widget.

Installation

Create a venv environment with Python 3.11 (recommended) or Python 3.10.

 python -m venv napari_easytrack-env

First, install napari.

Then, install easytrack via pip:

python -m pip install napari-easytrack

To install the latest development version of EpiTools clone this repository and run

python -m pip install -e .

Usage

To use napari-easytrack, first launch napari:

napari

Once in napari, click on the "Plugins" menu, then select "napari-easytrack" and click "Tracking" to open the tracking widget. We recommend starting with the Tracking widget to test the plugin with the provided presets.

Tracking Widget

Once in the Tracking widget, you can select one of the presets from the dropdown menu:

• Epithelial cells: for tracking epithelial cells in 2D+time datasets.
• Epithelial cells (Z-traacking): for tracking epithelial cells in 3D (space) datasets.
• Custom JSON: if none of the presets fit your data, you can provide a custom JSON file with tracking parameters optimised for your dataset. You can obtain this JSON file by first using the Parameter tuning widget.

Once you have selected your presets, select the "Segmentation Layer" to apply the tracking to and click "Apply Tracking". We also provide, in case it is needed, a "Clean Segmentation" and "Remove Small Objects" to improve the provided segmentation. In addition, you can also save your own configuration of parameters as a JSON file for future use by clicking on "Save Config (JSON)".

Parameter tuning Widget

To optimise your own tracking parameters specific to your dataset, you require to provide some ground-truth data with cells segmented and tracked. You will select this dataset as "Ground Truth Layer" in the Parameter tuning widget. As a first trial, we recommend using a small subset of your data (e.g., 10-20 frames) with a few cells tracked (e.g., 5-10 cells). With all the default parameters, click on "Start Optimization" to begin the optimisation process. You can cancel the process at any time by clicking on "Stop Optimization". Once the optimisation is finished, you can save the optimal parameters as a JSON file by clicking on "Save Config". You can then use this JSON file in the Tracking widget to track your entire dataset, selecting "Custom JSON" in the presets dropdown menu.

Issues

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