Image Reconstruction Algorithm for Modified SART Based on PM Model

Simultaneous Algebraic reconstruction technique (SART) has been widely applied to CT image reconstruction, but it is influenced by the noise interference, and as a result, the image quality is affected. In this article, we propose a modified SART algorithm to improve the quality of the image reconstruction, which constructs a new diffusion coefficient by the PM model, and makes the SART algorithm and PM model alternately according to the denoising principle of PM model. Experimental results demonstrate that the scheme can reduce the effect of noise on image quality substantially, increase the value of SNR and PSNR, and improve the performance of image reconstruction significantly.