Explainable radionuclide identification algorithm based on the convolutional neural network and class activation mapping
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Radionuclide identification is an important part of the nuclear material identification system. The development of artificial intelligence and machine learning has made nuclide identification rapid and automatic. However, many methods directly use existing deep learning models to analyze the gammaray spectrum, which lacks interpretability for researchers. This study proposes an explainable radionuclide identification algorithm based on the convolutional neural network and class activation mapping. This method shows the area of interest of the neural network on the gamma-ray spectrum by generating a class activation map. We analyzed the class activation map of the gamma-ray spectrum of different types, different gross counts, and different signal-to-noise ratios. The results show that the convolutional neural network attempted to learn the relationship between the input gamma-ray spectrum and the nuclide type, and could identify the nuclide based on the photoelectric peak and Compton edge. Furthermore, the results explain why the neural network could identify gamma-ray spectra with low counts and low signal-to-noise ratios. Thus, the findings improve researchers’ confidence in the ability of neural networks to identify nuclides and promote the application of artificial intelligence methods in the field of nuclide identification.
Abstract
Radionuclide identification is an important part of the nuclear material identification system. The development of artificial intelligence and machine learning has made nuclide identification rapid and automatic. However, many methods directly use existing deep learning models to analyze the gammaray spectrum, which lacks interpretability for researchers. This study proposes an explainable radionuclide identification algorithm based on the convolutional neural network and class activation mapping. This method shows the area of interest of the neural network on the gamma-ray spectrum by generating a class activation map. We analyzed the class activation map of the gamma-ray spectrum of different types, different gross counts, and different signal-to-noise ratios. The results show that the convolutional neural network attempted to learn the relationship between the input gamma-ray spectrum and the nuclide type, and could identify the nuclide based on the photoelectric peak and Compton edge. Furthermore, the results explain why the neural network could identify gamma-ray spectra with low counts and low signal-to-noise ratios. Thus, the findings improve researchers’ confidence in the ability of neural networks to identify nuclides and promote the application of artificial intelligence methods in the field of nuclide identification.
참고문헌 (30)
- Mario Gomez-Fernandez, "Status of research and development of learning-based approaches in nuclear science and engineering: A review", Nuclear Engineering and Design, 359, pp.110479~ (2020), Elsevier BV[DOI]
- Bon Tack Koo, "Development of a radionuclide identification algorithm based on a convolutional neural network for radiation portal monitoring system", Radiation Physics and Chemistry, 180, pp.109300~ (2021), Elsevier BV[DOI]
- J.T. Mihalczo, "NMIS plus gamma spectroscopy for attributes of HEU, PU and HE detection", Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 213, pp.378~384 (2004), Elsevier BV[DOI]
- Matt J. Aitkenhead, "Use of artificial neural networks in measuring characteristics of shielded plutonium for arms control", Journal of Analytical Atomic Spectrometry, 27(3), pp.432~ (2012), Royal Society of Chemistry (RSC)[DOI]
- Changfan Zhang, "Identification of SNM based on low-resolution gamma-ray characteristics and neural network", Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 927, pp.155~160 (2019), Elsevier BV[DOI]
- Mark Kamuda, "A comparison of machine learning methods for automated gamma-ray spectroscopy", Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 954, pp.161385~ (2020), Elsevier BV[DOI]
- Dajian Liang, "Rapid nuclide identification algorithm based on convolutional neural network", Annals of Nuclear Energy, 133, pp.483~490 (2019), Elsevier BV[DOI]
- S.M. Galib, "A comparative study of machine learning methods for automated identifi cation of radioisotopes using NaI gamma-ray spectra", Nuclear Engineering and Technology, 53(12), pp.4072~4079 (2021), 한국원자력학회[DOI]
- Sunci Wu, "Peak-searching method for low count rate <mml:math xmlns:mml="http:", 1998, pp.Math, MathML" display="inline" id="d1e597" altimg="si14.svg"><mml:mi>γ<[DOI]
- Sheng Qi, "Radionuclide identification method for NaI low-count gamma-ray spectra using artificial neural network", Nuclear Engineering and Technology, 54(1), pp.269~274 (2022), 한국원자력학회[DOI]
- Chunmiao Li, "A new radionuclide identification method for low-count energy spectra with multiple radionuclides", Applied Radiation and Isotopes, 185, pp.110219~ (2022), Elsevier BV[DOI]
- I. Rahwan, "Machine behaviour", Nature, 568, pp.477~486 (2019)[DOI]
- W. Samek, "Visualizing and Interpreting Deep Learning Models", pp.10~ (2017)
- G. Daniel, "Automatic and realtime identification of radionuclides in gamma-ray spectra : a new method based on convolutional neural network trained with synthetic data set", IEEE Trans. Nucl. Sci, 67, pp.644~653 (2020)[DOI]
- Mario Gomez-Fernandez, "Isotope identification using deep learning: An explanation", Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 988, pp.164925~ (2021), Elsevier BV[DOI]
- J. Ryu, "Development of neural network model with explainable AI for measuring uranium enrichment", IEEE Trans. Nucl. Sci, 68, pp.2670~2681 (2021)[DOI]
- S. Agostinelli, "Geant4—a simulation toolkit", Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 506(3), pp.250~303 (2003), Elsevier BV[DOI]
- "Technical and Functional Specifications for Border Monitoring Equipment" (2006)[URL]
- "Learning Multiple Layers of Features from Tiny Images" (2009)[URL]
- J. Snell, "Prototypical Networks for Few-Shot Learning", ~ (2017)
- M. Ren, "Meta-Learning for Semi-supervised Few-Shot Classification", ~ (2018)
- A. Krizhevsky, "ImageNet classification with deep convolutional neural networks", Commun. ACM, 60, pp.84~90 (2017)[DOI]
- A. Karpathy, "Largescale Video Classification with Convolutional Neural Networks", 8~ (2014)
- X. Zeng, "Gated Bi-directional CNN for object detection", Computer Vision e ECCV 2016, pp.354~369 (2016), Springer International Publishing
- "TensorFlow, Large-Scale Machine Learning on Heterogeneous Systems" (2015)[URL]
- C. Seifert, "Visualizations of deep neural networks in computer vision: a survey", Transparent Data Mining for Big and Small Data, pp.123~144 (2017), Springer International Publishing
- D. V. Carvalho, "Machine learning interpretability : a survey on methods and metrics", Electronics, 8, pp.832~ (2019)[DOI]
- "Toward Interpretable Machine Learning: Transparent Deep Neural Networks and beyond, CoRR. Abs/2003" (2020)[URL]
- B. Zhou, "Learning deep features for discriminative localization", 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.2921~2929 (2016), IEEE
- "People's Republic of China national metrology specification, Calibration specification for hand-held radiation monitors for detection and identification of radionuclides (JJF 1687-2018)", General Administration of Quality Supervision, Inspection and Quarantine, ~ (2018)