FUSE: Fluorescent Cell Labeling and Analysis

Specialized Pipeline for Cell Segmentation and ROI identification in Time-Series Data

The Berndt lab developed a cloud-based software called FUSE, designed to label and analyze fluorescent cells in time-series microscopy images. FUSE utilizes the Cellpose cell segmentation algorithm and a novel specialized cell labeling algorithm developed by the Berndt lab. With a user-friendly interface through Google Colab, FUSE allows users to efficiently analyze their data, providing a convenient, free, and fast method for analyzing timecourse data.

Table of Contents

• Motivations
• Features
• Prerequisites
• Customization
• Results and Output
• Keywords
• License
• Acknowledgments

Motivations

• Population analysis-based platform for fluorescence data
• Generates easy-to-use platform
• Aligns methods used to analyze fluorescent data
• Novel frame-by-frame cell identification algorithm that can be used with Cellpose segmentation
• Open source + Free platform
• Mitigating experimental burden associated with hand-drawn segmentation
• Cloud based storage and handling of data attenuates issues with data management and storage

Features

• Cloud-based implementation using Google Colab
• User-friendly interface with Google Colab forms
• Customizable pipeline for advanced users
• Integration with Cellpose for automated segmentation
• Convolutional autoencoder for feature extraction
• Analysis of fluorescence readouts (delta f) over time

Prerequisites

• Google Account for accessing Google Colab
• Cellpose-generated segmentation masks
• Fluorescence microscopy videos in TIF format

Getting Started

1. Clone or download the FUSE repository to your local machine.
2. Upload the repository to your Google Drive.
3. Open the main FUSE notebook in Google Colab.
4. Set up the Google Colab environment with the required dependencies.

Customization

Advanced users can modify the underlying code to better suit their specific cell types or experiment requirements. Customization options include adjusting the convolutional autoencoder parameters and modifying the tracking algorithm.

Results and Output

FUSE outputs the delta f (change in fluorescence) of the tracked cells over time, providing valuable insights into cell signaling events.

Keywords:

in vitro analysis algorithm, transparent data handling, high-throughput, unbiased, user-friendly, visual phenotyping

License

This project is licensed under the MIT License. See the LICENSE file for more details.

Copyright (c) 2023 University of Washington Department of Bioengineering

Acknowledgments

• Aida Moghadasi, Justin Daho Lee, Andre Berndt, PhD.
• University of Washington Department of Bioengineering, Mary Gates Research Endowment
• cellpose.org