Coronary System CT Image Denoising Study

Low dose computed tomography(LDCT)attracted many scholars and medical researchers in coronary artery medical imaging.However,the cardiac reconstruction image degradation in LDCT will affect the medical diagnostic accuracy.Therefore,acquiring better coronary artery image quality and reducing radiation dose as well has received much attention within the scientific community.This paper studies applications in coronary artery system LDCT by partial differential equation.A 4-neighborhood partial differential equation de-noising algorithm is proposed which is based on anisotropic diffusion equation.Regarding the ill posed problems of PM,the grads module of original image is taken place by the coronary artery sinogram which is smoothed by median filter.According to the control rate of relevant pixel values,novel PMG model has been proposed which defines a weight normalized gradient to get more effective at noise smoothing and coronary branches detail features preserving.Given that the coronary artery sinogram image change rule and the difficulty during choosing spread threshold,novel PMC model calculates spread threshold by mean absolute deviation and provides a time-varied diffusion function.The PMC model can be adaptive for the balance between noise removal and blood vessel edge preservation.PMGC model unifies PMG and PMC model to further improve the quality of coronary artery image reconstruction.Numerical results and CT scan data of the coronary artery reveal that the novel method can achieve better reconstruction image quality,which provides effective data for the clinical analysis.