The Development of a Feature-Driven Analytical Approach for Gamma-Ray Spectral Analysis

Authors

Aaron P. Fjeldsted, Penn State UniversityFollow
Jarek Glodo, Radiation Monitoring DevicesFollow
Darren Holland, Air Force Institute of TechnologyFollow
George V. Landon, Cedarville UniversityFollow
Clayton Scott, University of Michigan - Ann ArborFollow
Yilun Zhu, University of Michigan - Ann ArborFollow
Azaree T. Lintereur, Penn State UniversityFollow
Douglas E. Wolfe, Penn State UniversityFollow

Document Type

Article

Publication Date

7-2024

Abstract

Gamma-ray spectroscopy is an essential tool in nuclear science, nuclear security, and environmental monitoring. However, challenges arise in interpreting spectral data due to the presence of low counts, multiple sources, and dynamic backgrounds. To address these issues, a novel feature-driven analytical approach for gamma-ray spectral analysis using machine-learning techniques is developed. The method utilizes a series of random forest models for in-distribution (ID) multi-label classification, and the model-derived feature importance values to guide the out-of-distribution (OOD) detection task. The performance of this approach is quantitatively evaluated across various spectral parameters, including acquisition time, number of sources, energy of an OOD source, and background composition. Increasing the acquisition time from 1 s to 100 s leads to improved performance for multi-label classification, with 22 sources achieving F₁-scores ≥ 0.9 after 50 s acquisitions for a CLLBC handheld detector and a standoff distance of 30 cm. The feature-driven analytical approach also demonstrates robustness when handling complex source mixtures. Furthermore, it provides contextual energetic information for OOD detection. The results presented here highlight the interpretability of the approach, establishing clear links between the spectral features and underlying physics. Moreover, the approach effectively distinguishes overlapping spectral signatures of different ID gamma-ray sources, enhancing human reliability in machine learning-based gamma-ray spectral analysis. The feature-driven analytical approach offers a promising solution to automate gamma-ray spectral analysis by addressing existing limitations and providing insights into performance across diverse spectral parameters.

Comments

This is a subscription-access article, available to readers with a subscription to The Annals of Nuclear Energy, using the DOI link below.

Current AFIT students, faculty, and staff may access the full article by clicking here.

DOI

10.1016/j.anucene.2024.110464

Source Publication

Annals of Nuclear Energy

Recommended Citation

Fjeldsted, A. P., Glodo, J., Holland, D. E., Landon, G. V., Scott, C., Zhu, Y., Lintereur, A. T., & Wolfe, D. E. (2024). The Development of a Feature-Driven Analytical Approach for Gamma-Ray Spectral Analysis. Annals of Nuclear Energy, 202.