核事故致多核素内污染能谱成像测量技术研究

本项目针对核事故致多核素内污染的不均匀性、时变性、个体性，以及释放多射线的特点，通过研究宽能伽马射线与闪烁体作用位置定位、“位置-能谱”漂移修正、孔径编码能谱图像解码等关键问题研发一种新的多核素内污染能谱成像测量技术。首先，利用蒙特卡罗模拟与实验相结合的方法优化设计宽能近场伽马能谱成像探测器，降低散射，提高成像对比度、空间分辨率和能量分辨率。其次，建立256通道SiPM阵列信号同步读出系统；建立宽能伽马射线作用位置的卷积神经网络定位算法。然后，建立探测器网格化能量响应修正方法，消除探测器“位置-能谱”漂移；研究基于能谱分析的散射修正方法。最后，建立多核素内污染分布最大似然估计解码算法，获得多核素内污染分布图像。本研究对于减小个体内污染测量误差和内照射剂量估算误差具有十分重要的作用，有利于完善内照射剂量评估方法、提高我国辐射防护研究水平和核事故医学救治能力。

This project aims at the inhomogeneity, time-variability, individuality, and multi-ray emission of the multi-nuclide internal contamination caused by the radiological incident, we plan to develop a new spectral imaging measurement technology by solving a few key problems including the determination on the interaction position of the gamma ray on the scintillator, position-energy spectral shifting, MUAR coded aperture spectral imaging. Firstly, Monte Carlo simulation and experiments will be carried out to optimize detector for wide energy near-field imaging, lower the gamma scattering, improve the contrast, and resolution. Secondly, an asynchronous readout system for 256 SiPM array signals will be set up and the convolutional neural network algorithm for determining the interaction position of gamma ray on the scintillator will be designed. Thirdly, we will study a gridding energy correcting method to eliminate the detector’s dependence on energy response on the position. The scattering correction method will be studied based on the spectral analysis. Lastly, the maximum likelihood estimation method will be utilized to rebuild the internal contamination image. This research will help in reducing the error of measurement of internal contamination and evaluation of internal dose greatly, improving the method of internal dose evaluation and the ability of the radiation protection, promoting the medical treatment in the radiological incident.