Digital pulse processing algorithm for neutron and gamma rays discrimination

Hammad, M. E.; Kasban, H.; Fikry, R. M.; Dessouky, Moawad I.; Zahran, O.; Elaraby, S. M.; El‐Samie, Fathi E. Abd

doi:10.1007/s10470-019-01498-8

Cited by 9 publications

(5 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1) Classical Features: We use the term classical features to refer to features extracted from the raw pulses that are conventionally employed for classification i n s tandard lowrate settings. The classical features used in this paper are only a handful of the overall features that can be extracted to represent these pulses [10]- [12]. Although Challenge I makes feature extraction difficult under pileup conditions, some feature extraction methods are less prone to error than others.…”

Section: Methodsmentioning

confidence: 99%

“…Various PSD techniques were proposed in the literature. Traditional digital signal processing PSD methods include charge integration [9], slice-fitting algorithms [10], histogramdifference methods [11], and discrete wavelet transforms [12], [13]. PSD methods based on traditional machine learning techniques include Gaussian mixture models [14], support vector machines [15], and k-means [16].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Model-Based Deep Learning Algorithm for Pulse Shape Discrimination in High Event Rates

Morad,

Ghelman,

Ginzburg

et al. 2023

EPJ Web Conf.

View full text Add to dashboard Cite

Pulse shape discrimination (PSD) is at the core of radioactive particles monitoring. Conventional PSD methods are geared towards low event rates, and struggle in the presence of pileups resulting from high rate. In this work we develop a PSD algorithm that combines classic approaches with deep learning techniques, that is highly suitable for coping with the dramatic challenges associated with classifying pulses in high event rates. Common PSD algorithms for high event rates limit their research to two piled-up pulses. Our algorithm is designed and tested under severe pileup conditions, where three or more pulses were piled-up. We tested the algorithm on simulated data based on Cs2LiYCl6:Ce (CLYC) based detector pulse shapes and compare its performance to both traditional PSD algorithms and data-driven deep neural network (DNN) based algorithms. In high event rates, ranging up to 10 Mcps, the algorithm demonstrates up to 8 times fewer miss-classifications than the traditional normalized cross-correlation (NCC) approach, and up to 1.7 times fewer miss-classifications than a purely data-driven DNN-aided method.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Model-Based Deep Learning Algorithm for Pulse Shape Discrimination in High Event Rates

Morad,

Ghelman,

Ginzburg

et al. 2023

EPJ Web Conf.

View full text Add to dashboard Cite

show abstract

“…The proposed PSD methods are evaluated using simulated input signals generated using MATLAB environment. The mathematical fitting model presented in [18,19,20] is used for these generation. To observe the ability of the proposed approaches at lower and higher energies, the height of the simulated pulse is chosen to cover these energy bands [11,12].…”

Section: Data Descriptionmentioning

confidence: 99%

Efficient Discrimination of Neutron and Gamma Rays using Different Digital Pulse Processing Algorithms

Hammad

Kasban

Fikry

et al. 2020

Menoufia Journal of Electronic Engineering Research

Self Cite

View full text Add to dashboard Cite

In different neutron measurement experiments, it is necessary to apply pulse processing method to distinguish neutron pulses from gamma pulses. The discrimination process is based on the different decaying response of the detector for both neutron and gamma events. Different proposed algorithms are presented in this paper for determining the radiation type of detector output. In the proposed algorithms, features are extracted from the input radiation event. These features are extracted using charge integration, Hilbert transformation, and matched filtering methods. The extracted features are then fed the discriminator which is an Artificial Neural Network (ANN) or Support Vector Machine (SVM) discriminators. The obtained results prove that, the proposed approaches can be used efficiently for the neutron and gamma discrimination purpose and that method based on the Hilbert transformation achieves the highest discrimination rates.

show abstract

“…PSD methods realized in the analog and digital domain vary in their discrimination capabilities and power consumption. Traditional digital signal processing algorithms include slice-fitting, tail-sum, cosine similarity, histogramdifference, and Fourier analysis based methods [3]- [10]. In addition, recent advancements in the field of artificial intelligence have given rise to PSD techniques based on traditional machine learning [11]- [14] and neural networks [2], [15]- [20].…”

Section: Introductionmentioning

confidence: 99%

Analog Pulse Shape Discrimination Circuit for High Event Rates and Fast Scintillators with a Dynamic Deadtime

Morad,

Ghelman,

Osovizky

et al. 2023

EPJ Web Conf.

View full text Add to dashboard Cite

We developed an analog pulse shape discrimination (PSD) topology based on the well-established charge integration (CI) method, featuring two novel functional blocks beneficial for high event rate operation. The topology is designed for high-speed scintillators. The demonstrated analog design is potentially better suited than digital methods, when considering both processing time and power consumption aspects. The topology was tested using both experimental alpha and beta pulses from a plastic scintillator with a layer of ZnS(Ag) coupled to a PMT, and a fast digital emulator to simulate controlled high event rate scenarios. The discrimination capabilities of the topology were optimized and evaluated using a traditional figure of merit (FoM) approach. The topology achieved over 99% correct classifications when evaluated using the experimental pulses recorded. Additionally, the dedicated blocks resulted in a fourfold reduction in miss-classifications of slow pulses at an event rate of 100 kcps of fast pulses, while also providing a dynamic deadtime proportional to the pulse charge.

show abstract

Digital pulse processing algorithm for neutron and gamma rays discrimination

Cited by 9 publications

References 21 publications

Model-Based Deep Learning Algorithm for Pulse Shape Discrimination in High Event Rates

Model-Based Deep Learning Algorithm for Pulse Shape Discrimination in High Event Rates

Efficient Discrimination of Neutron and Gamma Rays using Different Digital Pulse Processing Algorithms

Analog Pulse Shape Discrimination Circuit for High Event Rates and Fast Scintillators with a Dynamic Deadtime

Contact Info

Product

Resources

About