Cargando…

Convolutional Neural Networks for Challenges in Automated Nuclide Identification

Improvements in Radio-Isotope IDentification (RIID) algorithms have seen a resurgence in interest with the increased accessibility of machine learning models. Convolutional Neural Network (CNN)-based models have been developed to identify arbitrary mixtures of unstable nuclides from gamma spectra. I...

Descripción completa

Detalles Bibliográficos
Autores principales: Turner, Anthony N., Wheldon, Carl, Wheldon, Tzany Kokalova, Gilbert, Mark R., Packer, Lee W., Burns, Jonathan, Freer, Martin
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8347253/
https://www.ncbi.nlm.nih.gov/pubmed/34372475
http://dx.doi.org/10.3390/s21155238
_version_ 1783735041757544448
author Turner, Anthony N.
Wheldon, Carl
Wheldon, Tzany Kokalova
Gilbert, Mark R.
Packer, Lee W.
Burns, Jonathan
Freer, Martin
author_facet Turner, Anthony N.
Wheldon, Carl
Wheldon, Tzany Kokalova
Gilbert, Mark R.
Packer, Lee W.
Burns, Jonathan
Freer, Martin
author_sort Turner, Anthony N.
collection PubMed
description Improvements in Radio-Isotope IDentification (RIID) algorithms have seen a resurgence in interest with the increased accessibility of machine learning models. Convolutional Neural Network (CNN)-based models have been developed to identify arbitrary mixtures of unstable nuclides from gamma spectra. In service of this, methods for the simulation and pre-processing of training data were also developed. The implementation of 1D multi-class, multi-label CNNs demonstrated good generalisation to real spectra with poor statistics and significant gain shifts. It is also shown that even basic CNN architectures prove reliable for RIID under the challenging conditions of heavy shielding and close source geometries, and may be extended to generalised solutions for pragmatic RIID.
format Online
Article
Text
id pubmed-8347253
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-83472532021-08-08 Convolutional Neural Networks for Challenges in Automated Nuclide Identification Turner, Anthony N. Wheldon, Carl Wheldon, Tzany Kokalova Gilbert, Mark R. Packer, Lee W. Burns, Jonathan Freer, Martin Sensors (Basel) Article Improvements in Radio-Isotope IDentification (RIID) algorithms have seen a resurgence in interest with the increased accessibility of machine learning models. Convolutional Neural Network (CNN)-based models have been developed to identify arbitrary mixtures of unstable nuclides from gamma spectra. In service of this, methods for the simulation and pre-processing of training data were also developed. The implementation of 1D multi-class, multi-label CNNs demonstrated good generalisation to real spectra with poor statistics and significant gain shifts. It is also shown that even basic CNN architectures prove reliable for RIID under the challenging conditions of heavy shielding and close source geometries, and may be extended to generalised solutions for pragmatic RIID. MDPI 2021-08-03 /pmc/articles/PMC8347253/ /pubmed/34372475 http://dx.doi.org/10.3390/s21155238 Text en © 2021 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Turner, Anthony N.
Wheldon, Carl
Wheldon, Tzany Kokalova
Gilbert, Mark R.
Packer, Lee W.
Burns, Jonathan
Freer, Martin
Convolutional Neural Networks for Challenges in Automated Nuclide Identification
title Convolutional Neural Networks for Challenges in Automated Nuclide Identification
title_full Convolutional Neural Networks for Challenges in Automated Nuclide Identification
title_fullStr Convolutional Neural Networks for Challenges in Automated Nuclide Identification
title_full_unstemmed Convolutional Neural Networks for Challenges in Automated Nuclide Identification
title_short Convolutional Neural Networks for Challenges in Automated Nuclide Identification
title_sort convolutional neural networks for challenges in automated nuclide identification
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8347253/
https://www.ncbi.nlm.nih.gov/pubmed/34372475
http://dx.doi.org/10.3390/s21155238
work_keys_str_mv AT turneranthonyn convolutionalneuralnetworksforchallengesinautomatednuclideidentification
AT wheldoncarl convolutionalneuralnetworksforchallengesinautomatednuclideidentification
AT wheldontzanykokalova convolutionalneuralnetworksforchallengesinautomatednuclideidentification
AT gilbertmarkr convolutionalneuralnetworksforchallengesinautomatednuclideidentification
AT packerleew convolutionalneuralnetworksforchallengesinautomatednuclideidentification
AT burnsjonathan convolutionalneuralnetworksforchallengesinautomatednuclideidentification
AT freermartin convolutionalneuralnetworksforchallengesinautomatednuclideidentification