An Improved Two-Shot Tracking Algorithm for Dynamics Analysis of Natural Killer Cells in Tumor Contexts.

Area of Science:

• Immunology
• Biophysics
• Computational Biology

Background:

• Natural killer cells (NKCs) are crucial for innate immunity and anticancer immunotherapy.
• Understanding NKC motion and collective dynamics is vital for developing effective cancer treatments.
• Current tracking algorithms struggle with NKC analysis due to their small size and complex environments.

Purpose of the Study:

• To analyze NKC morphology and key points using deep learning.
• To improve multi-object tracking algorithms for NKC dynamics.
• To explore factors influencing NKC motion and their interaction with cancer cells.

Main Methods:

• Deep learning for NKC morphology analysis.
• Development of Distance Cascade Matching and Re-Search tracking methods.
• Analysis of over 5000 image frames (approx. 300,000 cells) to study NKC dynamics.

Main Results:

• Polarized NKCs exhibit enhanced activity and directed movement along their polarization axis.
• Polarization promotes NKC approach towards cancer cells.
• NKC distribution on cancer cell surfaces follows a Boltzmann distribution at equilibrium.

Conclusions:

• Optimized tracking algorithms enhance the study of NKC dynamics.
• Cell morphology, particularly polarization, significantly influences NKC behavior and cancer cell interaction.
• NKC collective behavior can be modeled, providing insights into immunotherapy mechanisms.