skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Automated isotope identification algorithm using artificial neural networks

Abstract

There is a need to develop an algorithm that can determine the relative activities of radio-isotopes in a large dataset of low-resolution gamma-ray spectra that contains a mixture of many radio-isotopes. Low-resolution gamma-ray spectra that contain mixtures of radio-isotopes often exhibit feature over-lap, requiring algorithms that can analyze these features when overlap occurs. While machine learning and pattern recognition algorithms have shown promise for the problem of radio-isotope identification, their ability to identify and quantify mixtures of radio-isotopes has not been studied. Because machine learning algorithms use abstract features of the spectrum, such as the shape of overlapping peaks and Compton continuum, they are a natural choice for analyzing radio-isotope mixtures. An artificial neural network (ANN) has be trained to calculate the relative activities of 32 radio-isotopes in a spectrum. Furthermore, the ANN is trained with simulated gamma-ray spectra, allowing easy expansion of the library of target radio-isotopes. In this paper we present our initial algorithms based on an ANN and evaluate them against a series measured and simulated spectra.

Authors:
ORCiD logo [1];  [1]; ORCiD logo [1]
  1. Univ. of Illinois at Urbana-Champaign, Urbana, IL (United States)
Publication Date:
Research Org.:
Univ. of Michigan, Ann Arbor, MI (United States)
Sponsoring Org.:
USDOE National Nuclear Security Administration (NNSA), Office of Nonproliferation and Verification Research and Development (NA-22)
Contributing Org.:
Department of Nuclear, Plasma, and Radiological Engineering, University of Illinois at Urbana-Champaign, Urbana, Il, 61801 USA.
OSTI Identifier:
1367506
Grant/Contract Number:
NA0002534
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
IEEE Transactions on Nuclear Science
Additional Journal Information:
Journal Volume: 64; Journal Issue: 7; Journal ID: ISSN 0018-9499
Publisher:
IEEE
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; 73 NUCLEAR PHYSICS AND RADIATION PHYSICS; artificial neural network; automated isotope identification; gamma-ray spectroscopy; machine learning

Citation Formats

Kamuda, Mark, Stinnett, Jacob, and Sullivan, Clair. Automated isotope identification algorithm using artificial neural networks. United States: N. p., 2017. Web. doi:10.1109/TNS.2017.2693152.
Kamuda, Mark, Stinnett, Jacob, & Sullivan, Clair. Automated isotope identification algorithm using artificial neural networks. United States. doi:10.1109/TNS.2017.2693152.
Kamuda, Mark, Stinnett, Jacob, and Sullivan, Clair. Wed . "Automated isotope identification algorithm using artificial neural networks". United States. doi:10.1109/TNS.2017.2693152. https://www.osti.gov/servlets/purl/1367506.
@article{osti_1367506,
title = {Automated isotope identification algorithm using artificial neural networks},
author = {Kamuda, Mark and Stinnett, Jacob and Sullivan, Clair},
abstractNote = {There is a need to develop an algorithm that can determine the relative activities of radio-isotopes in a large dataset of low-resolution gamma-ray spectra that contains a mixture of many radio-isotopes. Low-resolution gamma-ray spectra that contain mixtures of radio-isotopes often exhibit feature over-lap, requiring algorithms that can analyze these features when overlap occurs. While machine learning and pattern recognition algorithms have shown promise for the problem of radio-isotope identification, their ability to identify and quantify mixtures of radio-isotopes has not been studied. Because machine learning algorithms use abstract features of the spectrum, such as the shape of overlapping peaks and Compton continuum, they are a natural choice for analyzing radio-isotope mixtures. An artificial neural network (ANN) has be trained to calculate the relative activities of 32 radio-isotopes in a spectrum. Furthermore, the ANN is trained with simulated gamma-ray spectra, allowing easy expansion of the library of target radio-isotopes. In this paper we present our initial algorithms based on an ANN and evaluate them against a series measured and simulated spectra.},
doi = {10.1109/TNS.2017.2693152},
journal = {IEEE Transactions on Nuclear Science},
number = 7,
volume = 64,
place = {United States},
year = {Wed Apr 12 00:00:00 EDT 2017},
month = {Wed Apr 12 00:00:00 EDT 2017}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:
  • An artificial neural network was designed and used to develop a dose algorithm for a multi-element thermoluminescence dosimeter (TLD). The neural network architecture is based on the concept of functional links network (FLN). Neural network is an information processing method inspired by the biological nervous system. A dose algorithm based on neural networks is fundamentally different as compared to conventional algorithms, as it has the capability to learn from its own experience. The neural network algorithm is shown the expected dose values (output) associated with given responses of a multi-element dosimeter (input) many times. The algorithm, being trained that way,more » eventually is capable to produce its own unique solution to similar (but not exactly the same) dose calculation problems. For personal dosimetry, the output consists of the desired dose components: deep dose, shallow dose and eye dose. The input consists of the TL data obtained from the readout of a multi-element dosimeter. The neural network approach was applied to the Harshaw Type 8825 TLD, and was shown to significantly improve the performance of this dosimeter, well within the U.S. accreditation requirements for personnel dosimeters.« less
  • This paper proposes the use of Artificial Neural Networks (ANN`s) to identify and control an induction machine. Two systems are presented: a system to adaptively control the stator currents via identification of the electrical dynamics, and a system to adaptively control the rotor speed via identification of the mechanical and current-fed system dynamics. Various advantages of these control schemes over other conventional schemes are cited and the performance of the combined speed and current control scheme is compared with that of the standard vector control scheme.
  • Power system loads are important for planning and operation of an electric power system. Load characteristics can significantly influence the results of synchronous stability and voltage stability studies. This paper presents a methodology for identification of power system load dynamics using neural networks. Input-output data of a power system dynamic load is used to design a neural network model which comprises delayed inputs and feedback connections. The developed neural network model can predict the future power system dynamic load behavior for arbitrary inputs. In particular, a third-order induction motor load neural network model is developed to verify the methodology. Neuralmore » network simulation results are illustrated and compared with the induction motor load response.« less
  • In the present work, we have performed analyte species and concentration identification using an array of ten differentially functionalized microcantilevers coupled with a back-propagation artificial neural network pattern recognition algorithm. The array consists of ten nanostructured silicon microcantilevers functionalized by polymeric and gas chromatography phases and macrocyclic receptors as spatially dense, differentially responding sensing layers for identification and quantitation of individual analyte(s) and their binary mixtures. The array response (i.e. cantilever bending) to analyte vapor was measured by an optical readout scheme and the responses were recorded for a selection of individual analytes as well as several binary mixtures. Anmore » artificial neural network (ANN) was designed and trained to recognize not only the individual analytes and binary mixtures, but also to determine the concentration of individual components in a mixture. To the best of our knowledge, ANNs have not been applied to microcantilever array responses previously to determine concentrations of individual analytes. The trained ANN correctly identified the eleven test analyte(s) as individual components, most with probabilities greater than 97%, whereas it did not misidentify an unknown (untrained) analyte. Demonstrated unique aspects of this work include an ability to measure binary mixtures and provide both qualitative (identification) and quantitative (concentration) information with array-ANN-based sensor methodologies.« less
  • In this paper we present a multi-frequency excitation eddy current differential transducer and dynamic neural models which were used to detect and identify artificial flaws in thin conducting plates. Plates are made of Inconel600. EDM notches have relative depth from 10% to 80% and length from 2 mm to 7 mm. All flaws were located on the opposite surface of the examined specimen. Measured signals were used as input for training and verifying dynamic neural networks with a moving window. Wide range of ANN (Artificial Neural Network) structures are examined for different window length and different number of frequency componentsmore » in excitation signal. Observed trends are presented in this paper.« less