CAJ | 학술논문

采用Geant4模拟软件模拟不同形状γ源的MURA（19×19）嵌套型编码成像，分别利用两种相关解码算法（精细采样平衡相关和δ解码）和最大似然-期望最大化迭代算法对各个编码像进行数值重建，用对比度-噪声比（CNR）对各种算法的重建图像质量进行比较。结果表明，三种图像重建算法均能为各种不同形状的平面γ源给出清晰、准确的图像，其中MLEM算法重建图像的分辨率和CNR比传统的相关解码要高，该算法可以较好地抑制统计噪声，图像质量高。因此，编码孔成像技术用于γ相机成像是可行的。
채용Geant4모의연건모의불동형상γ원적MURA（19×19）감투형편마성상，분별이용량충상관해마산법（정세채양평형상관화δ해마）화최대사연-기망최대화질대산법대각개편마상진행수치중건，용대비도-조성비（CNR）대각충산법적중건도상질량진행비교。결과표명，삼충도상중건산법균능위각충불동형상적평면γ원급출청석、준학적도상，기중MLEM산법중건도상적분변솔화CNR비전통적상관해마요고，해산법가이교호지억제통계조성，도상질량고。인차，편마공성상기술용우γ상궤성상시가행적。
Coded aperture imaging of various shaped γsources was simulated by Geant4 toolkit, with a mosa-icked MURA(19 ×19)mask.Two correlation decoding algorithms (Finely Sampled Balanced and δdecoding), as well as the iterative Maximum Likelihood -Expectation Maximization (MLEM) algorithm, were applied for digital reconstruction of the coded images .The qualities of reconstructed images were compared by calculating the contract-to-noise ratio (CNR).The results showed that each of the algorithms represented a clear , exact image for every single source and that high -quality images with a better resolution and CNR were obtained by MLEM algorithm, as the statistic noise was effectively inhibited .Accordingly, coded aperture imaging was ap-plicable in γcamera.