LivecellX

LivecellX is a deep-learning-based, single-cell object-oriented framework designed for quantitative analysis of single-cell dynamics in long-term, label-free live-cell imaging datasets.

Manuscript

For detailed methods and validations, please see our manuscript:

> Ni et al., LivecellX: A Deep-learning-based, Single-Cell Object-Oriented Framework for Quantitative Analysis in Live-Cell Imaging (__stub__, under review).

Key Features

  • Segmentation & Tracking: Accurate deep learning-based segmentation, integrated tracking algorithms (SORT, b-track).

  • Corrective Segmentation Network (CS-Net): Automatically correct over- and under-segmentation errors using context-aware deep learning models.

  • Trajectory-Level Correction: Algorithms utilizing temporal consistency for error correction and accurate lineage reconstruction.

  • Biological Process Detection: Automated detection and classification of cellular processes (mitosis, apoptosis).

  • High-dimensional Feature Extraction: Including morphological (Active Shape Models), textural (Haralick, LBP), and deep learning-based features (VAE).

  • Object-Oriented Data Structure: Intuitive and efficient management of single-cell trajectories and features, enabling multi-dataset integration.

  • Napari GUI Integration: Interactive visualization, manual correction, and lineage tracing.

  • Parallelized Computation: Efficient processing of large datasets using multi-core computation.

Getting Started

Indices and Tables