Nuclear Imaging for Treaty Verifica5on with Enhanced Informa5on Security

Nuclear imaging systems have been proposed as a technology component for treaty verificaFon between countries. Imaging systems can acquire data that contain both spaFal informaFon about the object being probed and spectral informaFon regarding the material composiFon of the object. Thus it is expected that accurate verificaFon of treaty-accountable items (TAIs) can be achieved using imaging data. A drawback to imaging techniques is that the monitor can use the detector data and knowledge of the imaging system to reconstruct sensiFve geometrical informaFon regarding the treaty- accountable item (TAI) being imaged. Thus, it is convenFonally thought that a physical or so_ware informaFon barrier (IB) must be used to prevent the disclosure of sensiFve informaFon to the monitor. We have developed methods for uFlizing imaging data for treaty-verificaFon imaging while miFgaFng the need for an IB. These methods all rely on the concept of list-mode processing where individual neutron and gamma-ray events are processed immediately a_er detecFon and never stored. Thus, an image is never actually formed; instead, a single test staFsFc is updated as events come in. This test staFsFc is used to make the final decision as to whether the item being imaged is a TAI or not. The methods used to process this list-mode data, which we call observer models, are designed to be insensiFve to specific aspects of the objects being imaged. Thus, even if the monitor has access to the methods being used to make decisions, they would not be able to discover informaFon about the TAI that the host wants to keep hidden. In this paper, we will present an overview of treaty-verificaFon imaging without informaFon barriers. We will discuss our methods in general terms, how these methods might be uFlized in a treaty-verificaFon campaign, and the work sFll needed to achieve the goal of nuclear imaging in treaty verificaFon without an informaFon barrier.