A NEURAL-NETWORK METHOD FOR COUNTING AND LOCATING OVERLAPPING SCINTILLATION PULSES IN DIGITIZED WAVEFORMS
DOI:
https://doi.org/10.46991/PYSUA.2026.60.2.102Keywords:
pulse pile-up, scintillation detectors, neural networks, waveform reconstruction, convolutional networks, template subtractionAbstract
At high counting rates, scintillation pulses overlap in the digitizer window, and when the arrival times are not fixed by an accelerator clock, the classical reconstruction methods begin to lose them. We present a neural-network method that counts the overlapping pulses and measures the arrival time and amplitude of each. The method is studied on simulated $1~GHz$ waveforms containing zero to six pulses. Two small convolutional networks recovered $99.2\%$ of the individual pulses, with a mean timing error of $4.6~ns$ and an amplitude error of $1.1$ ADC counts ($9\%$ of a typical amplitude); the pulse count was determined correctly in $96.5\%$ of the windows. Under identical conditions and scoring, a matched filter recovered only $46\%$ of the pulses with a $7.5~ns$ timing error, and an iterative CLEAN-type template subtraction $57\%$ with a $5.2~ns$ error. The false-positive rate was $0.4\%$ for the network, $2.6\%$ for CLEAN, and $12\%$ for the matched filter.
Downloads
References
Knoll G.F. Radiation Detection and Measurement (4th ed.). Wiley (2010).
Spieler H. Semiconductor Detector Systems. Oxford University Press (2005). DOI: https://doi.org/10.1093/acprof:oso/9780198527848.001.0001
IAEA. Instrumentation for Digital Nuclear Spectroscopy. IAEA-TECDOC-1706. International Atomic Energy Agency (2013).
Lee M., Lee D., et al. Pulse Pileup Correction Method for Gamma-ray Spectroscopy in High Radiation Fields. Nucl. Eng. Technol. 52 (5) (2020), 1029-1035. https://doi.org/10.1016/j.net.2019.12.003 DOI: https://doi.org/10.1016/j.net.2019.12.003
Cleland W.E., Stern E.G. Signal Processing Considerations for Liquid Ionization Calorimeters in a High Rate Environment. Nucl. Instrum. Meth. A 338 (2-3) (1994), 467-497. https://doi.org/10.1016/0168-9002(94)91332-3 DOI: https://doi.org/10.1016/0168-9002(94)91332-3
CMS Collaboration. Reconstruction of Signal Amplitudes in the CMS Electromagnetic Calorimeter in the Presence of Overlapping Proton-proton Interactions. JINST 15 (2020), P10002. https://doi.org/10.1088/1748-0221/15/10/P10002 DOI: https://doi.org/10.1088/1748-0221/15/10/P10002
H"ogbom J.A. Aperture Synthesis with a Non-regular Distribution of Interferometer Baselines. Astron. Astrophys. Suppl. 15 (1974), 417-426.
Guo W., Gardner R.P., Mayo C.~W. A Study of the Real-time Deconvolution of Digitized Waveforms with Pulse Pile up for Digital Radiation Spectroscopy. Nucl. Instrum. Meth. A 544 (3) (2005), 668-678. https://doi.org/10.1016/j.nima.2004.12.036 DOI: https://doi.org/10.1016/j.nima.2004.12.036
Kuhn H.W. The Hungarian Method for the Assignment Problem. Naval Research Logistics Quarterly 2 (1955), 83-97. DOI: https://doi.org/10.1002/nav.3800020109
Yu D., Kolbaek M., et al. Permutation Invariant Training of Deep Models for Speaker-independent Multi-talker Speech Separation. Proc. ICASSP (2017). https://doi.org/10.1109/ICASSP.2017.7952154 DOI: https://doi.org/10.1109/ICASSP.2017.7952154
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Proceedings of the YSU

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.