Computer simulation uses theoretical models to predict the performance of a real system. Computer simulation can be divided into two general categories: analytical and statistical. Analytically based simulation generates system models using known analytical equations. A good example of this category is the generation of projections of mathematical phantoms. This type of projection is calculated based on the line integrals of attenuation coefficients of objects whose shapes can be described by closed-form equations (the phantom is formed with cylinders, spheres, ellipsoids, or bars). For a given source and detector cell position, the line integral can be calculated precisely for each object. The final result is the weighted summation of these integrals. Statistically based simulation, on the other hand, uses random number generators and the physical property of the interaction process to predict the system performance. A good example is the Monte Carlo simulation used to predict the scatter distribution.
Online access to SPIE eBooks is limited to subscribing institutions.