Nuclide Identification using CsI(Tl) Gamma Ray Spectra and Neural Networks

Maiwald T, Leder E, Pijahn R, Buchhold R, Fischer G (2023)

Publication Language: English

Publication Type: Conference contribution, Conference Contribution

Publication year: 2023

Pages Range: 1624-1628

Conference Proceedings Title: 21st IEEE International Conference on Machine Learning and Applications (ICMLA)

Event location: Nassau, The Bahamas BS

DOI: 10.1109/ICMLA55696.2022.00249

Abstract

With increasing popularity of machine learning methods used in gamma ray spectroscopy for radioisotope identification, quality and scope of the data, to train and evaluate models, constitute the key towards applicable approaches. This paper demonstrates several steps needed to be considered in order to create a novel dataset, which is motivated by physical, technical and standardization requirements. Therefore offset shifts, gain drifts, variating acquisition times and nuclide mixtures are considered. Furthermore, initial classification results gained from a neural network approach are presented.

Authors with CRIS profile

Timo Maiwald Lehrstuhl für Technische Elektronik (LTE) Georg Fischer Professur für Technische Elektronik

Involved external institutions

Thermo Fisher Scientific Messtechnik GmbH DE Germany (DE)

How to cite

APA:

Maiwald, T., Leder, E., Pijahn, R., Buchhold, R., & Fischer, G. (2023). Nuclide Identification using CsI(Tl) Gamma Ray Spectra and Neural Networks. In 21st IEEE International Conference on Machine Learning and Applications (ICMLA) (pp. 1624-1628). Nassau, The Bahamas, BS.

MLA:

Maiwald, Timo, et al. "Nuclide Identification using CsI(Tl) Gamma Ray Spectra and Neural Networks." Proceedings of the International Conference on Machine Learning and Applications 2022, Nassau, The Bahamas 2023. 1624-1628.

BibTeX: Download