超重三核素的核乳胶数据结合机器学习的超氚核束缚能精确测量

Aug 5, 2026

A machine learning workflow is developed to identify production and decay events of hypertriton in nuclear emulsion images for high-precision determination of its binding energy. The approach builds on established object-detection techniques and is trained with surrogate images produced via Monte Carlo simulations and image-transfer methods. The method has already detected the first hypertriton candidate using only 10−4 of the full emulsion dataset, indicating that sufficient statistics will soon enable a precise hypertriton binding-energy measurement.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Introduction

Nuclear emulsion technique and background challenge

Motivation: renewed interest in light hypernuclei and hypertriton

Suplemento de la Revista Mexicana de Fsica 3 0308122 (2022) 1–6 Precise measurement on the binding energy of hypertriton from the nuclear emulsion data using analysis with machine learning A. Kasagia;b , E. Liua;c;d , M. Nakagawaa , H. Ekawaa , J. Yoshidaa;e , W. Doua;f , A. Muneema;g , K. Nakazawab;h , C. Rappoldi , N. Saitoa , T. R. Saitoa;j;k , M. Takil , Y. K. Tanakaa , and H. Wanga a High Energy Nuclear Physics Laboratory, Cluster for Pioneering Research, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan b Graduate School of Engineering, Gifu University, 1-1 Yanagido, Gifu 501-1193, Japan c Institute of Modern Physics, Chinese Academy of Sciences, 509 Nanchang Road, Lanzhou, Gansu Province, 730000, China. d School of Nuclear Science and Technology, University of Chinese Academy of Sciences, No. 19(A) Yuquan Road, Shijingshan District, Beijing, 100049, China. e Department of physics, Tohoku University, Aramaki, Aoba-ku, Sendai 980-8578, Japan. f Department of Physics, Saitama University, Saitama, 338-8570, Japan. g Faculty of Engineering Sciences, Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi, 23640, KP, Pakistan h Faculty of Education, Gifu University, 1-1 Yanagido, Gifu 501-1193, Japan. i Instituto de Estructura de la Materia, CSIC, Madrid, Spain. j GSI Helmholtz Centre for Heavy Ion Research, Planckstrasse 1, D-64291 Darmstadt, Germany. k School of Nuclear Science and Technology, Lanzhou University, 222 South Tianshui Road, Lanzhou, Gansu Province, 730000, China. l Graduate School of Artiﬁcial Intelligence and Science, Rikkyo University, 3-34-1 Nishi Ikebukuro, Toshima-ku, Tokyo 171-8501, Japan. Received 15 January 2022; accepted 11 February 2022 A machine learning model has been developed to search for events of production and decay of a hypertriton in nuclear emulsion data, which is used for measuring the binding energy of the hypertriton at the best precision. The developed model employs an established technique for object detection and is trained with surrogate images generated by Monte Carlo simulations and image transfer techniques. The ﬁrst hypertriton event has already been detected with the developed method only with 10 ¡4 of the total emulsion data. It implies that a sufﬁcient number of hypertriton events will soon be detected for the precise measurement of the hypertriton binding energy. Keywords: Hypernucleus; hypertriton; binding energy; nuclear emulsion; machine learning. DOI: [https://doi.org/10.31349/SuplRevMexFis.3.0308122](https://doi.org/10.31349/SuplRevMexFis.3.0308122) 1. Introduction Hypernuclei have been studied for approximately seven decades to investigate the baryonic interactions under the ﬂavored SU(3) symmetry. The ﬁrst hypernucleus was observed in nuclear emulsions, which were irradiated by high-energy cosmic rays [1] . After the discovery, the binding energies of the ¤-hyperon, i.e., the ¤ separation energy, were studied with the nuclear emulsion technique and K ¡ meson beams at CERN [2, 3] . Recently, experimental techniques with realtime particle counters using secondary meson-and primary electron-beams as well as heavy ion beams have been established and are employed for studying the binding energies, structure, and lifetimes of hypernuclei [4–6] . However, the emulsion technique still has a great advantage in measuring the binding energy of light hypernuclei at the best precision. A nuclear emulsion is a special photographic ﬁlm, which is able to detect the tracks of charged particles. When a charged particle passes through an emulsion layer, its threedimensional trajectory is recorded and can be visualized after chemical developments. Thanks to the small size of silver halide crystal, approximately 200 nm, a spatial resolution of better than a micrometer for reconstructing tracks of charged particles can be achieved by observing clusters of grains with an optical microscope. For hypernuclear studie

FAQs

为什么使用核乳胶技术测量超氚核束缚能？

核乳胶能记录带电粒子的三维轨迹，通过测量衰变相关粒子的动能与动量推导不变质量，从而确定束缚能；同时具有极高的空间分辨率，利于对轻超核进行高精度测量。

核乳胶数据中束缚能测量面临的主要困难是什么？

生产与衰变超核相关的事件轨迹与大量无关背景轨迹在乳胶中叠加且不含时间信息，必须从巨大背景中挖掘出有效的超核事件。

该文提出的机器学习方法取得了怎样的早期结果？

开发的目标检测模型利用蒙特卡洛生成的替代图像与图像迁移技术进行训练，并已在仅使用总数据的10−4时检测到首个超氚核事件，表明后续可获得足够统计量用于精确测量。

Aug 5, 2026

Upload by Lucas Martin