Cell-ACDC

Cell-ACDC performs segmentation, tracking, and cell cycle annotation of live-cell imaging data to enable quantitative analysis of single-cell dynamics.

Key Features:

• Deep learning segmentation: Incorporates two deep learning models for single-cell segmentation tailored for yeast and mammalian cells and implemented with TensorFlow and PyTorch.
• Cell tracking: Includes a cell tracking method for continuous monitoring and analysis of cellular dynamics over time.
• Semi-automated label-free cell cycle annotation: Provides a semi-automated workflow for annotating cell cycle states without additional labeling techniques.
• Python implementation: Implemented in Python.
• Modular design: Modularized to allow integration of additional segmentation and downstream image-analysis methods.

Scientific Applications:

• Single-cell quantitative analysis: Enables detailed quantitative analysis of cellular processes at the single-cell level.
• High-throughput microscopy and microfluidics: Handles high-throughput imaging data generated from microfluidic platforms.
• Dynamic cellular events: Supports studies of cell division, migration, and responses to stimuli through segmentation and tracking.
• Pedigree and signal analysis: Facilitates pedigree analysis and signal quantification derived from segmentation and tracking outputs.

Methodology:

Implemented in Python using two deep learning models for single-cell segmentation (yeast and mammalian) implemented with TensorFlow and PyTorch, together with a cell tracking method and a semi-automated label-free cell cycle annotation workflow.