The simulation of a data acquisition system for a proposed high resolution PET scanner

Abstract The simulation of a specific data acquisition (DAQ) system architecture for a proposed high resolution Positron Emission Tomography (PET) scanner is discussed. Stochastic processes are used extensively to model PET scanner signal timing and probable DAQ circuit limitations. Certain architectural parameters, along with stochastic parameters, are varied to quantitatively study the resulting output under various conditions. The inclusion of the DAQ in the model represents a novel method of more complete simulations of tomograph designs, and could prove to be of pivotal importance in the optimization of such designs. I. INTRODUCTION The DAQ system under study was designed for possible use with a proposed PET scanner for the University of Chicago. Design of the scanner and DAQ is presently on hold due to funding uncertainties and the desire to explore recent technological developments. This paper will focus only on the simulation of the primary DAQ system architecture. Although several other aspects of the DAQ system including control, diagnostics, attenuation scan filtering, and accidental rate measuements were briefly considered, none of them will be addressed in this paper. Soon after the DAQ architecture was proposed, it became apparent that due to the inherent stochastic processes. a simulation was needed to study its behavior. The simulation is written using the high level Verilog hardware descriptive language using mostly behavioral models rather than stmchual constructs. This enables easy modifications to DAQ characteristics within the context of probable hardware limitations without the restraint of dealing with a specific hardware configuration. However, the ultimate goal of the design process is to replace these behavioral models with models of “real” parts in order to migrate toward pc board design and simulation in a Cadence CAE environment.