Generalization ability of a CNNγ-ray localization model for radiation imaging
作者机构:Chengdu University of TechnologyChengdu 610059China Sichuan University of Science and EngineeringZigong 643000China
出 版 物:《Nuclear Science and Techniques》 (核技术(英文))
年 卷 期:2023年第34卷第12期
页 面:53-65页
核心收录:
学科分类:082704[工学-辐射防护及环境保护] 08[工学] 080203[工学-机械设计及理论] 0827[工学-核科学与技术] 0802[工学-机械工程] 0702[理学-物理学]
基 金:supported by the National Natural Science Foundation of China(Nos.41874121 and U19A2086)
主 题:γ-Ray imaging γ-Ray localization model Convolutional neural network Spatial resolution
摘 要:Inγ-ray imaging,localization of theγ-ray interaction in the scintillator is *** neural network(CNN)techniques are highly promising for improvingγ-ray *** study evaluated the generalization capabilities of a CNN localization model with respect to theγ-ray energy and thickness of the *** model maintained a high positional linearity(PL)and spatial resolution for ray energies between 59 and 1460 *** PL at the incident surface of the detector was 0.99,and the resolution of the central incident point source ranged between 0.52 and 1.19 *** modified uniform redundant array(MURA)imaging systems using a thick crystal,the CNNγ-ray localization model significantly improved the useful field-of-view(UFOV)from 60.32 to 93.44%compared to the classical centroid localization ***,the signal-to-noise ratio of the reconstructed images increased from 0.95 to 5.63.