基于SART算法的CL硬化伪影校正方法研究

Research of Beam Hardening Correction Method for CL System Based on SART Algorithm

X射线薄板层析成像(CL)系统是一种针对板状构件的新型断层成像系统,该系统中存在射束硬化现象。本工作研究了一种基于SART算法的硬化伪影校正算法(SART-BHC算法),通过对该算法的研究及对CL系统矩阵的计算,使该方法能应用于CL中。通过模拟与实验相结合的方式,利用蒙特卡罗方法获得CL投影的模拟数据,在实际CL中获得实验数据,然后利用SART-BHC与FPB算法对两种数据分别重建。结果表明,SART-BHC算法可应用于CL中,不需要任何先验知识,能很好地校正硬化伪影,并对层间混叠有抑制作用。

Computed laminography （CL） is a non-destructive testing technique for large objects ,especially for planar objects .Beam hardening artifacts were wildly observed in the CL system and significantly reduce the image quality .This study proposed a novel simultaneous algebraic reconstruction technique （SART ） based beam hardening correction （BHC） method for the CL system ,namely the SART-BHC algorithm in short .The SART-BHC algorithm took the polychromatic attenuation process in account to formu-late the iterative reconstruction update .A novel projection matrix calculation method which was different from the conventional cone-beam or fan-beam geometry was also studied for the CL system .The proposed method was evaluated with simulation dataand experimental data ,which was generated using the Monte Carlo simulation toolkit Geant4 and a bench-top CL system ,respectively .All projection data were reconstructed with SART-BHC algorithm and the standard filtered back projection （FBP） algorithm . The reconstructed images show that beam hardening artifacts are greatly reduced with the SART-BHC algorithm compared to the FBP algorithm .The SART-BHC algorithm doesn’t need any prior know-ledge about the object or the X-ray spectrum and it can also mitigate the interlayer aliasing .