CytoCensus

CytoCensus performs automated three-dimensional (3D) cell detection and quantification in complex multilayered tissues using supervised machine learning to enable analysis of cell numbers, distributions, and divisions.

Key Features:

• 3D Detection Capability: Identifies and quantifies cells within volumetric/3D datasets, including multilayered tissues.
• Supervised Machine Learning: Uses supervised learning to classify and detect cells from annotated training data.
• 3D Point-and-Click Training: Extends 2D point-and-click annotation into three-dimensional space for model training.
• Robustness to Ill-Defined Boundaries: Detects cells in datasets with poorly defined cell boundaries.
• Comparative Performance: Demonstrated superior accuracy and speed of cell detection in comparative tests against other freely available image analysis software.

Scientific Applications:

• Stem Cell Quantification: Counting stem cells and their progeny to analyze cellular dynamics.
• Time-Lapse Analysis of Drosophila Larval Brains: Quantification of individual cell divisions from time-lapse movies of explanted Drosophila larval brains to compare wild-type and mutant phenotypes.
• Zebrafish Retinal Organoids: Analysis of 3D organization of multiple cell classes within zebrafish retinal organoids.
• Mouse Embryo Cell Distributions: Examination and quantification of cell distributions in mouse embryos.

Methodology:

Training a supervised machine learning model using user-provided annotations via a point-and-click approach extended into 3D to enable automated cell detection in volumetric datasets.