Signal Discrimination in Thinned Silicon Neutron Detectors using Machine learning

Anderson, Mike and Prendergast, David and Alhamdi, Mustafa and Cheneler, David and Monk, Stephen (2019) Signal Discrimination in Thinned Silicon Neutron Detectors using Machine learning. In: 2019 IEEE Nuclear Science Symposium and Medical Imaging Conference, 2019-10-26 - 2019-11-02, Manchester Central Convention Centre.

Text (MA IEEE Poster 221019 V2)
MA_IEEE_Poster_221019_V2.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial.
Download (1MB)

Abstract

High gamma backgrounds can pose a significant source of interference in solid-state neutron detectors making the neutron flux approximation inaccurate. This work focuses on optimizing a thin sensor thickness to enhance the neutron capture rate and reject gammas, and analysis of multiple input source through the differentiation of signals using pattern recognition. Gamma isotopes and neutron spectrums have been simulated using GEANT4 + Electronic noise estimation. Different machine learning tools have been considered to discriminate different gamma and neutron sources, including PCA, RNN, SVM, KNN, ResNet and others.