CT图像重建算法对比研究

Comparative study of CT image reconstruction algorithms

对4种经典的重建算法:滤波反投影算法(FBP)、代数重建算法(ART)、联合代数重建算法(SART)、基于最小化总变差的联合代数重建算法(SART-TV)进行对比研究。利用Shepp-Logan头模型进行仿真实验,在稀疏角度和有限角度情况下,对比重建效果。仿真结果表明:在2种数据情况下,FBP算法重建出的图像效果比其他3种算法差,但其重建所需的时间最短;ART算法在2种数据情况下,重建出的图像效果比FBP好,但是随着迭代次数的增加,耗时比FBP长;SART算法获得的重建图像比ART好,但是耗时比ART长;SART-TV算法在本文2种数据情况下,获得的重建图像比其他3种算法更好。

A comparative study is conducted on four classical reconstruction algorithms,which are filter back-projection algorithm(FBP),algebraic reconstruction techniques algorithm(ART),simultaneous algebraic reconstruction technique algorithm(SART),and simultaneous algebraic reconstruction technique algorithm based on minimizing the total variation(SART-TV).By using Shepp-Logan head model for simulation test,the reconstruction images were compared when the angle was sparse and limited.The simulation result shows that in the two data cases,the reconstructed image obtained by the FBP algorithm is worse than the other 3 algorithms,but it takes the shortest time;the images which are reconstructed by the ART algorithm in the two data cases is better than FBP,but with the increase in the number of iterations,it takes longer than FBP;the reconstructed images obtained by the SART algorithm is better than those of ART,but it takes longer than ART;the reconstructed images of SART-TV algorithm under the two data conditions in this paper is better than the other 3 algorithms.