Modeling and Tracking of Heterogeneous Cell Populations via Open Multi-Agent Systems.

Area of Science:

• Biomedical Research
• Cellular Dynamics
• Optical Microscopy

Background:

• Understanding live-cell behaviors in vitro is crucial for biomedical research.
• Optical microscopy is a key technique for observing cellular dynamics.
• Tracking dynamic changes in cell populations, including mitosis and migration, remains challenging, especially in complex co-cultures.

Purpose of the Study:

• To introduce an enhanced cell-tracking algorithm for analyzing dynamic changes in heterogeneous cell populations within co-culture models.
• To accurately model and predict cell movements, interactions, and proliferation in complex cellular environments.
• To validate the algorithm using a novel dataset of tumor and normal cell interactions.

Main Methods:

• Modeling cell movements and interactions using tailored open multi-agent systems for co-culture experiments.
• Parameter identification using real data for a multi-agent, multi-culture framework.
• Embedding the model within an Extended Kalman Filter to predict heterogeneous cell population dynamics across video frames.

Main Results:

• The enhanced algorithm effectively tracks heterogeneous cell types, including osteosarcoma and mesenchymal stromal cells.
• The method accurately captures cell-cell interactions and proliferation dynamics in a challenging co-culture model.
• Performance metrics demonstrated superior effectiveness compared to state-of-the-art methodologies, including the generation of estimated cell lineage trees.

Conclusions:

• The developed algorithm significantly advances cell-tracking capabilities for complex co-culture models.
• This tool provides a robust framework for studying cancer cell evolution and interactions with stromal cells.
• The algorithm's ability to predict cellular dynamics and generate lineage trees offers valuable insights for biomedical research.