A High-Performance Cellular Automaton Model of Tumor Growth with Dynamically Growing Domains

Area of Science:

• Computational Biology
• Mathematical Modeling
• Oncology

Background:

• Simulating tumor growth from a single cell to a macroscopic mass involves vast spatial and temporal scales.
• Traditional cellular automata models for tumor growth face computational performance limitations.
• Efficient simulation requires optimized data structures, memory management, and domain setup.

Purpose of the Study:

• To propose a high-performance cellular automaton model for simulating tumor growth.
• To enhance computational efficiency for multi-scale Monte Carlo simulations of tumor development.
• To analyze parameter sensitivity and non-monotonic dependencies in tumor volume.

Main Methods:

• Developed a cellular automaton model with a dynamically expanding domain.
• Implemented optimized memory access, data structures, and cell handling techniques.
• Utilized multi-scale Monte Carlo simulations for tumor growth analysis.

Main Results:

• Achieved high-performance computation for multi-scale tumor growth simulations.
• Identified tumor properties that benefit the proposed high-performance design.
• Demonstrated non-monotonic relationships between tumor volume and various parameters.

Conclusions:

• The proposed cellular automaton model offers efficient and scalable tumor growth simulation.
• Dynamic domain expansion and optimized computational techniques are key to performance.
• Understanding parameter sensitivity is crucial for accurate tumor volume prediction.