Explosives detection studies using Fast-Neutron Transmission Spectroscopy

Fast-Neutron Transmission Spectroscopy (FNTS) is being investigated for detection of explosives in luggage or air cargo. We present here the principle results of a two-year study of a few-view tomographic FNTS system using the Monte Carlo radiation transport code MCNP to simulate neutron transmission through simple luggage phantoms and Receiver Operator Characteristic (ROC) curves to determine system performance. Elemental distributions along projections through the interrogated object are obtained by analyzing MCNP generated neutron transmission data. Transmission data for few (3-5) angles and relatively coarse resolution ({approximately}2 cm) are used to create a tomographic reconstruction of elemental distributions within the object. The elemental unfolding and tomographic reconstruction algorithms and the concept of transmission-derived cross sections for use in elemental analysis have been validated by application to experimental data. Elemental distributions are combined in an explosives detection algorithm to provide an indication of the presence or absence of explosives. The algorithm in current use, termed the ``equivalent explosive`` algorithm, determines the quantity of explosive that can be formed using the measured amount of the constituent elements in each pixel. Reconstruction and explosives detection algorithms have been applied to a series of randomly packed suitcases to generated ROC that describe system performance inmore » terms of the probability of detection and of false alarms. System studies have been performed to study the operational characteristics and limitations of a FNTS system, and to determine the system`s sensitivity to several important parameters such as neutron source reaction and incident particle energy, flight path length, and the position of the interrogated object.« less