SYNTHETIC SCORING SCHEME FOR ISOTOPE IDENTIFICATION IN SMARTID

摘要

Gamma Spectroscopy is widely used to identify radioactive nuclides. It also plays an important role in nuclear safeguards and non-proliferation applications for Special Nuclear Material (SNM) detection, characterization, and identification. A new rapid gamma spectroscopy post processing algorithm, SmartIDXP, or ‘SmartID’, synthetically enhances the resolution of low resolution scintillator detectors with an accurate, adaptive nuclide identification and attribution. The spectra post-processed by SmartID is synthetically (numerically) enhanced comparable with higher resolution detectors, improving, for example, sodium or cesium iodide resolving power by as much as a factor of three, with adaptive analysis to attribute shielded nuclides. In this study, a new scoring scheme is developed to accurately identify different radioactive nuclides supported by a comprehensive radionuclide database. The database includes more than 2300 known photopeaks from more than 220 gamma-emitting nuclides. The novel scoring scheme evaluates a matching score for each nuclide in the database based on photopeaks extracted by SmartID from a raw input spectrum. The nuclides with the highest score(s) are considered to be the most likely radiation sources for the given spectrum. The scoring scheme enhances SmartID’s capability to identify radiation sources. We tested the performance of the synthetic scoring scheme on a set of spectra acquired using a standard 2”×2” sodium iodide detector on various sources and shielding environments. The results show that the synthetic scoring scheme can accurately identify the sources for more ～95% of all spectra tested.