Addressing spectral variability with spectral unmixing for radionuclides automatic detection and quantification in the context of embedded measurements

Read full text Summarise with VIRAT
Authors: Charles-Philippe Mano, Hugues Paradis, Luc Patryl, Sébastien Gutierrez, Nicolas Comte, F. Carrel
Cited by: 0
References: 13

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

In recent years, the development of spectral unmixing methods incorporating Poisson counting statistics has attracted sustained attention in gamma-ray spectrometry. These approaches have significantly improved sensitivity in low-countingstatistics conditions and have demonstrated robustness for both radionuclide identification and quantification. However, because they rely on libraries of reference spectra, they may remain sensitive to environment-induced spectral variations, which can degrade performance.This work presents a new spectral unmixing method that accounts for spectral variability. Two new algorithms incorporating spectral variability are described and then tested on both simulated and experimental data, in embedded measurement configurations where spectral variability is significant. The validity of these algorithms is demonstrated both through the reconstruction of spectral signatures from different initial signatures and in terms of sensitivity.

Fields of Study & Subjects

Radiation Detection and Scintillator Technologies Radioactive Decay and Measurement Techniques Radioactivity and Radon Measurements

Publication details

Journal / source
HAL
Publication date
1 January 2026
Language
EN
Document type
Preprint