阵列式康普顿相机GEANT4模拟与图像重建及优化

出于对放射性核素各种应用和安全的考虑，需要性能良好的康普顿相机。通过蒙特卡罗方法模拟了双层阵列式康普顿相机，散射探测器材料为Si，吸收探测器材料为CZT（碲锌镉）。利用反投影图像重建算法进行了图像重建，采用极大似然法对重建结果进行优化，有效地提高了重建图像的分辨率。当放射源与探测器距离100 mm时，经过10次极大似然法迭代计算，位置分辨率与角分辨率分别达到5.1 mm与7.3。检验了极大似然法对于康普顿相机放射源图像重建的优化作用。

GEANT4 simulation and image reconstruction with optimization for array type Compton camera

For various applications and security considerations of radioactive nuclide, good performance of Compton camera is very important. In this article, the Monte Carlo method is used to simulate the double array type Compton camera, whose scattering detector material is Si and absorption detector material is CZT. By using the projection image reconstruction algorithm for image reconstruction, the maximum likelihood method is adopted to optimize reconstruction results, which effectively improves the resolution of the reconstructed image. When the source-detector distance is 100 mm, after 10 times of maximum likelihood method iterative calculation, the position resolution and angle resolution are 5.1 mm and 7.3 respectively. It is proved that maximum likelihood method can optimize Compton camera source image reconstruction effectively.