摘要
时序编码成像图像重建受噪声影响,导致技术应用受限。为解决这一问题,通过分析时序编码投影成像基本原理,构建源射线分布重建图像质量评价指标:对比度噪声比(CNR)和均方根误差(RMSE);设计M-P双正码与M-P正—反码两种编码板,采用MCNP模拟计算两种编码板对^(252)Cf进行时序编码成像的投影像相对强度,由MLEM算法进行图像重建获得射线源分布;通过分析两种编码模式重建图像CNR和RMSE受不同强度白噪声的影响,表明M-P正—反码编码板比M-P双正码重建图像的CNR高、RMSE低;并对两种编码模式重建图像CNR和RMSE受FOV场外点源伪影的影响进行分析,表明M-P正—反码编码板比M-P双正码重建图像的CNR高、RMSE高,认为正—反码编码模式在抑制噪声、检测视场外射线源伪影、提高重建图像质量方面具备一定优势,时序编码成像的编码板设计时可优先选择正反码编码模式。
The quality of reconstruction image for time-encoded imaging(TEI)is affected by noise,which limits its application.To solve this problem,the principle of time-encoded projection imaging was illustrated,contrast-to-noise ratio(CNR)and root mean square error(RMSE)was introduced to evaluate the quality of reconstruction image.Then,the rotating masks were designed by M-P mask-mask and M-P mask-anti-mask,and the MCNP was used to simulate TEI process of two masks for ^(252)Cf surface source,and to obtain the relative projection intensity at different time series.Thus,reconstruction image of the radiation source distribution can be obtained by reconstructing the projection using the MLEM algorithm.Through analyzing the impact of white noise with different intensity on the value of CNR and RMSE for reconstruction image by two masks,the results show that the values of CNR for reconstruction image by M-P mask-anti-mask is higher than that of M-P mask-mask,the RMSE is exactly reverse.And the impact of the artifacts on the CNR and RMSE of reconstruction image was analyzed,it indicated that the CNR and RMSE of reconstruction image by M-P mask-anti-mask are both higher than that of M-P mask-mask.The results show that the mask-anti-mask mode has certain advantages in suppressing white noise,detecting artifacts of ray sources outside the FOV,and improving the quality of reconstruction image,it can be preferred in the design of the rotating mask for TEI.
作者
汤适荣
张松柏
伍晓东
黄孟
朱剑钰
TANG Shirong;ZHANG Songbai;WU Xiaodong;HUANG Meng;ZHU Jianyu(School of Automation and Information Engineering,Sichuan University of Science and Engineering,Zigong 643002,China;School of Physics and Electronic Engineering,Sichuan University of Science and Engineering,Zigong 643002,China;Centre for Strategic Studies,China Academy of Engineering Physics,Beijing 100088,China)
出处
《核电子学与探测技术》
CAS
北大核心
2023年第5期955-963,共9页
Nuclear Electronics & Detection Technology
基金
四川省科技厅项目资助(资助号:2021YFSY0058)。