Optimal experimentation for nuclear medicine imaging system design

Abstract

We investigated the potential applications of the response surface methodology (RSM) for nuclear medicine imaging systems optimization. RSM is a technique used to improve system or process design in a wide range of applications in engineering. RSM is essentially a technique and body of knowledge that helps find the set of design parameters that will achieve the best system performance by conducting a minimal number of experiments. It also helps us determine the effect of parameters and their interactions on the system. While traditional approaches to experimentation consider the effect of each parameter separately, RSM relies on the simultaneous optimization with respect to all parameters. This technique is particularly suitable for nuclear medicine imaging systems since the cost of real or simulated experiments is very high and therefore a systematic and efficient experimentation with a simultaneous parameter optimization scheme is essential. RSM can also be used for multiple objective optimization problems where more than one system performance variable is optimized. In this paper, we first present the foundations of RSM and its applications in various fields. We then give an example from a breast scintigraphy collimator optimization problem to illustrate the use of RSM in nuclear medicine imaging systems optimization.