List-mode source injection algorithm for detectors with arbitrary pose and trajectory

Abstract

When evaluating detectors and algorithms for nuclear threat detection in populated environments, introducing actual sources is usually neither feasible nor economical. It is more practical to move the detector, (which is either airborne, vehicle-borne, or human portable) through the test environment and then to later artificially superimpose either measured or simulated source signatures on the recorded background data. We present a source injection algorithm that can be used to inject either measured or simulated source data into list-mode background data. It is designed to accommodate cases where measured “injection data” are available only at a limited set of locations. We describe a sampling scheme suitable for obtaining measured source injection data. We then use these data to demonstrate the source injection algorithm applied to list-mode data collected with a helicopter-borne detector system, where arbitrary detector poses and trajectories are possible. Stochastic methods are used both to select source events from a dataset containing both source and background events, and to scale the number of selected events to match the field conditions of detector-source distance, air attenuation, acquisition duration, and the relative strength of the measured and injected sources. This algorithm is planned to be used as part of the Airborne Radiation Enhanced-sensor System (ARES) Advanced Technology Demonstration.