Pulse shape discrimination and exploration of scintillation signals using convolutional neural networks

Abstract

We demonstrate the use of a convolutional neural network to perform neutron-gamma pulse shape discrimination, where the only inputs to the network are the raw digitised silicon photomultiplier signals from a dual scintillator detector element made of 6Li F:ZnS(Ag) scintillator and PVT plastic. A realistic labelled dataset was created to train the network by exposing the detector to an AmBe source, and a data-driven method utilising a separate photomultiplier tube was used to assign labels to the recorded signals. This approach is compared to the charge integration and continuous wavelet transform methods and a simpler artificial neural net. It is found to provide superior levels of discrimination, achieving an area under the curve of 0.996 ± 0.003. We find that the neural network is capable of extracting interpretable features directly from the raw data. In addition, by visualising the high-dimensional representations of the network with the t-SNE algorithm, we discover that not only is this method robust to minor mislabeling of the training dataset but that it is possible to identify an underlying substructure within the signals that goes beyond the original labelling. This technique could be utilised to explore and cluster complex, raw detector data in a novel way that may reveal more insights than standard analysis methods.

Pulse Shape Discrimination of low-energy nuclear and electron recoils for improved particle identification in NaI:Tl

“…These approaches can then be extended to other machine learning techniques, such as Convolutional Neural Networks (CNNs), to extract the features embedded in each waveform to better discriminate between nuclear and electron recoil events. CNNs have seen use in discrimination between neutron and gamma pulse shapes previously [21], but this approach is reliant on greater statistics for network training.…”

Timing and characterization of shaped pulses with MHz ADCs in a detector system: a comparative study and deep learning approach

“…Recently, deep learning [13] as a renewed machine learning technique has progressed rapidly. It has been successfully used for particle/event discrimination and identification at the pulse level [14], the pixel level [15] and the voxel (three-dimensional) level [16]. In view of the fact that neural networks are applicable to classification tasks as well as regression tasks, it is meaningful to explore the capability of deep learning in the above-mentioned pulse timing problem.…”

A machine learning-based methodology for pulse classification in dual-phase xenon time projection chambers

“…Pulses are identified by the classification algorithms based on their characteristic geometrical features. Some analyses use the raw shape of the pulse directly to identify the pulse (e.g., using convolutional neural networks [28] or deep learning [29]) but typically an intermediate parameterization step is used to calculate relevant shape-related quantities, such as integrated areas, lengths, fit parameters and other detector-specific traits, that are then used by the classification algorithms. This work will follow the latter approach.…”