Final Progress Report: Isotope Identification Algorithm for Rapid and Accurate Determination of Radioisotopes Feasibility Study

Rawool-Sullivan, Mohini¹  Bounds, John Alan¹  Brumby, Steven P.¹  Prasad, Lakshman¹  Sullivan, John P.¹ 

[1] Los Alamos National Laboratory

关键词: 07 ISOTOPES AND RADIATION SOURCES;    ALGORITHMS;    BREMSSTRAHLUNG;    ESCAPE PEAKS;    FEASIBILITY STUDIES;    IDENTIFICATION SYSTEMS;    ISOTOPES;    LEARNING;    NEUTRONS;    PROGRESS REPORT;    RADIATIONS;    RADIOISOTOPES;    SENSORS;    SHIELDING;   

DOI  :  10.2172/1039677
RP-ID  :  LA-UR-12-20990
PID  :  OSTI ID: 1039677
Others  :  TRN: US1202270

美国|英语

来源: SciTech Connect

【 摘 要 】

This is the final report of the project titled, 'Isotope Identification Algorithm for Rapid and Accurate Determination of Radioisotopes,' PMIS project number LA10-HUMANID-PD03. The goal of the work was to demonstrate principles of emulating a human analysis approach towards the data collected using radiation isotope identification devices (RIIDs). It summarizes work performed over the FY10 time period. The goal of the work was to demonstrate principles of emulating a human analysis approach towards the data collected using radiation isotope identification devices (RIIDs). Human analysts begin analyzing a spectrum based on features in the spectrum - lines and shapes that are present in a given spectrum. The proposed work was to carry out a feasibility study that will pick out all gamma ray peaks and other features such as Compton edges, bremsstrahlung, presence/absence of shielding and presence of neutrons and escape peaks. Ultimately success of this feasibility study will allow us to collectively explain identified features and form a realistic scenario that produced a given spectrum in the future. We wanted to develop and demonstrate machine learning algorithms that will qualitatively enhance the automated identification capabilities of portable radiological sensors that are currently being used in the field.