Coupled minimum-cost flow cell tracking.

Area of Science:

• Cell biology
• Bioimage analysis
• Computational biology

Background:

• Automated cell monitoring is crucial for high-throughput screening and high-content analysis.
• Accurate cell tracking must account for diverse cellular behaviors such as mitosis, occlusion, and movement.
• Existing methods may struggle with the complexity of real-time cell population dynamics.

Purpose of the Study:

• To develop and validate a novel cell tracking algorithm for automated monitoring in complex biological screening.
• To explicitly model and accurately detect various cell behaviors, including mitosis and merging events.
• To provide a robust solution for analyzing large-scale cell population dynamics.

Main Methods:

• Utilized a graph-theoretic minimum-cost flow framework to model cell tracking.
• Extended the minimum-cost flow algorithm to incorporate mitosis and merging events by coupling specific graph edges.
• Applied the algorithm to a large dataset comprising nearly 6,000 images, 400,000 cells, and 32,000 tracks across multiple experimental datasets.

Main Results:

• The developed algorithm demonstrated high accuracy in tracking individual cells across diverse behaviors.
• Achieved an accuracy exceeding 99% for cell tracking and behavior detection in large-scale datasets.
• Successfully handled complex scenarios including mitosis, occlusion, rapid movement, and cells entering/leaving the field of view.

Conclusions:

• The proposed graph-theoretic approach provides a highly accurate and robust method for automated cell tracking in high-content screening.
• The algorithm's ability to model complex cell behaviors enhances the reliability of large-scale biological data analysis.
• This advancement facilitates more precise monitoring of cell populations in various screening applications.