Performance analysis of a computational off-lattice tumor growth model

Abstract

This paper presents a performance analysis of a three dimensional off-lattice agent based model of tumor growth. The model uses short range potential functions to handle mechanical interactions. This computationally intensive process can be accelerated by various reduction techniques. We focus on two different methods, a spatial decomposition which is extremely useful for simulating low density suspended cell cultures, and a heuristic to reduce computations for simulating solid tumors. We show that these techniques can significantly improve simulation performance in appropriate situations.